Dyslexia Statistics 2026: Prevalence, School, Assistive Tech

Dyslexia statistics depend on where researchers draw the line. In a worldwide meta-analysis of primary school children, studies that set a strict reading cutoff found dyslexia in 5.36% of children at 1.5 standard deviations below the mean and 5.32% at 2 standard deviations, against 9.10% in studies that did not report a cutoff (Brain Sciences, Yang et al., Feb 2022). A model built from meta-analytic data found that samples of poor readers contain more children whose low reading is expected from their other skills than children with the unexpected gap that defines dyslexia (Journal of Learning Disabilities, Wagner et al., May 2020). The gap between estimates comes from the definition: a strict test cutoff counts far fewer children than a loose one.

Schools measure who qualifies for support. In fall 2024, 2,516,690 students aged 5 to 21 in the US and outlying areas qualified for special education under specific learning disability, the IDEA category that includes dyslexia, which is 33% of the 7,590,772 school-age students served under the law (U.S. Department of Education, Office of Special Education Programs, IDEA Section 618 Child Count and Educational Environments 2024-25, Dec 2025). England's schools recorded 180,217 pupils whose primary special need is a specific learning difficulty in 2025/26, 2.2% of pupils in state-funded schools and 10.5% of all pupils with special educational needs (Department for Education, Special educational needs in England, Jun 2026).

The statistics on dyslexia below cover prevalence, genetics, school results, special education, adult outcomes and assistive technology, and each links to the organisation that produced it.

Key Takeaways

About 7 in 100 primary school children have developmental dyslexia: 7.10%, pooled across 58 studies worldwide (Brain Sciences, Yang et al., Feb 2022).
8.9% of US children aged 3 to 17 had been told they have a learning disability in 2025, up from 6.6% in 2019 (CDC National Center for Health Statistics, NHIS, Aug 2026).
English pupils whose primary need is a specific learning difficulty received 21.0 suspensions per 100 pupils in 2024/25, nearly double the 10.88 rate for all pupils (Department for Education, Suspensions and permanent exclusions in England, Jul 2026).
45% of children with a parent or sibling with reading problems develop dyslexia, against 11.6% of children without that family history (Psychological Bulletin, Snowling & Melby-Lervag, Jan 2016).
36.5% of US children with ADHD also have a learning disability (Journal of Clinical Child & Adolescent Psychology, CDC and HRSA, May 2024).
72% of US fourth-graders with disabilities scored below NAEP Basic in reading in 2024, against 34% of fourth-graders without a disability (NCES, The Nation's Report Card, Jan 2025).
Only 32% of English 11-year-olds whose main need is a specific learning difficulty met the expected writing standard in 2026, against 86% of pupils with no special need (Department for Education, Key stage 2 attainment, Sep 2026).
UK Access to Work approvals for people with dyslexia rose from 1,390 in 2007/08 to a peak of 6,310 in 2023/24, then fell to 4,840 in 2024/25 (Department for Work and Pensions, Access to Work statistics April 2007 to March 2025, Jan 2026).
Specific learning disability covers 46.5% of school-age special education students in New Mexico but 18.7% in Kentucky (U.S. Department of Education, Office of Special Education Programs, IDEA Section 618 Child Count and Educational Environments 2024-25, Dec 2025).
84.8% of students with a specific learning disability in the US and outlying areas who left school in 2023-24 graduated with a regular diploma, against 77.9% of all students with disabilities (U.S. Department of Education, Office of Special Education Programs, IDEA Section 618 Exiting 2023-24, Sep 2025).
The share of US adults at the lowest literacy level rose from 19% in 2017 to 28% in 2023 (NCES, PIAAC 2023 U.S. National Results, Dec 2024).
Only 24% of US college students with learning disabilities had told their school about it, and 17% received accommodations (IES, National Longitudinal Transition Study-2, Sep 2011).
61% of employers said the accommodation their employee needed cost nothing, and the median one-time cost was $300 (Job Accommodation Network, Costs and Benefits of Accommodation, Sep 2025).
Text-to-speech and read-aloud tools raised reading comprehension for students with reading difficulties by an average effect size of d = .35 (Journal of Learning Disabilities, Wood et al., Jan 2017).

How common is dyslexia around the world?

Dyslexia prevalence statistics disagree mainly because they use different cutoffs. Reading skill runs on a continuum, so the share labelled dyslexic depends on where a study or a school system draws the line.

Dyslexia statistics worldwide: pooled prevalence of developmental dyslexia in primary school children split by sex, writing system, orthographic depth, diagnostic cutoff, grade and study size, from a meta-analysis of 58 studies and 168,216 children — Figure 1: A meta-analysis of 58 studies (56 articles) of primary school children shows why no single number holds: the strictest cutoffs and the largest samples sit well below the pooled figure, while alphabetic and logographic scripts barely differ. Source: Brain Sciences, Yang et al., Prevalence of Developmental Dyslexia in Primary School Children: A Systematic Review and Meta-Analysis, Feb 2022. Chart published by the source, captured 2 Oct 2026.

It depends on the test. The stricter the reading cutoff and the larger the sample, the smaller the share of children a study classes as dyslexic.

Bigger studies found less dyslexia too: studies of more than 10,000 children averaged 3.13%, against 8.43% in studies of 500 to 1,000 children (Brain Sciences, Yang et al., Feb 2022).
The writing system made no significant difference: prevalence was 7.26% among children reading alphabetic scripts and 6.97% among children reading logographic scripts such as Chinese (Brain Sciences, Yang et al., Feb 2022).
Spelling consistency did not change the rate either: 7.13% in languages with shallow orthographies, where letters map reliably to sounds, against 7.55% in deep orthographies (Brain Sciences, Yang et al., Feb 2022).

A low reading score is not the same as dyslexia, and a label given at age 6 does not always hold.

In a simulation built from meta-analytic data, only 30% of readers 1.5 standard deviations below the mean had reading comprehension 1.5 standard deviations or more below their listening comprehension, the unexpected gap that defines dyslexia in most research (Journal of Learning Disabilities, Wagner et al., May 2020).
In a Connecticut cohort of 414 children, only 7 of the 25 children (28%) classified as dyslexic in grade 1 were still classified that way in grade 3, the result a normal-distribution model of reading predicts (New England Journal of Medicine, Shaywitz et al., Jan 1992).
Among more than 1.1 million 23andMe research participants, a self-reported dyslexia diagnosis was more common in younger adults (5.34% of 20- to 30-year-olds) than older ones (3.23% of 80- to 90-year-olds), a gap consistent with rising diagnosis over time (Nature Genetics, Doust et al., Oct 2022).

Regional meta-analyses show the same spread.

Across 18 studies of 30,243 children aged 6 to 13 in Arab countries, pooled dyslexia prevalence was 11%, and it was twice as high in Arab countries in Asia (24%) as in Africa (12%) (Research in Developmental Disabilities, Aldakhil, Jul 2024).
In Iran, 18 studies of 28,952 primary school children put dyslexia at 5.7%, but the test used moved the result from 3.4% with informal reading tests to 8.8% with a screening inventory (Iranian Journal of Psychiatry, Sakhai et al., Apr 2025).

How many children in the US and UK have dyslexia or a learning disability?

Neither country counts dyslexia directly. The US tracks parent-reported learning disabilities, and England, Scotland and Wales each record school needs in a different way, so their numbers sit side by side and are never added together.

Family income splits the US rate almost in half: 13.0% of children below the federal poverty line had a learning disability in 2025, against 7.1% in families at 200% of the poverty line or more (CDC National Center for Health Statistics, NHIS, Aug 2026).
Before the recent rise, the US rate was flat for nearly a decade: 7.74% of children aged 3 to 17 had a learning disability across 2009 to 2017, and the rate was 7.86% in both 2009 to 2011 and 2015 to 2017 (Pediatrics, Zablotsky et al., Oct 2019).
The number of English pupils recorded with a specific learning difficulty as their primary need grew 19% in a decade, from 151,559 in 2015/16, while total pupil numbers grew 4.8% (Department for Education, Special educational needs in England, Jun 2026).
Most of those pupils get school-level help rather than a statutory plan: 88.1% of English pupils with a specific learning difficulty were on SEN support in 2025/26, and 11.9% had an education, health and care plan (Department for Education, Special educational needs in England, Jun 2026).
Identification in England climbs after primary school: 110,098 secondary pupils had a specific learning difficulty as their primary need in 2025/26, against 65,001 primary pupils (Department for Education, Special educational needs in England, Jun 2026).

The other UK nations count differently, and the gap between their figures is mostly about method.

Scotland, which records every reason a pupil needs support rather than one primary need, counted 39,036 pupils with dyslexia in 2025, 5.6% of its 695,923 pupils (Scottish Government, Pupil census supplementary statistics, Dec 2025).
Wales recorded 1,710 reports of dyslexia as an additional learning need in maintained schools in January 2026, 0.4% of pupils, after its move to a new ALN system (Welsh Government, Schools' census results: January 2026, Aug 2026).
Which school a child attends changes the odds of being identified: in a study of about 540,000 Year 6 pupils in England, the median odds ratio between two schools was close to three even after adjusting for pupil and school characteristics (Journal of Learning Disabilities, Daniel et al., Feb 2026).

Is dyslexia more common in boys than girls?

Yes, in most studies, but the gap is smaller when researchers test every child than when schools decide who gets referred.

In a Connecticut study, schools identified 13.6% of second-grade boys and 3.2% of girls as reading disabled, while testing every child found 8.7% of boys and 6.9% of girls (JAMA, Shaywitz et al., Aug 1990).
In the worldwide meta-analysis, 9.22% of primary school boys had dyslexia against 4.66% of girls (Brain Sciences, Yang et al., Feb 2022).
Four epidemiological studies in New Zealand and the UK found reading disability in 17.6% to 21.6% of boys against 7.9% to 13.0% of girls, with odds ratios from 1.39 to 3.19 (JAMA, Rutter et al., Apr 2004).
The gap widens with severity: among 491,103 Florida second-graders, the boy-to-girl ratio peaked at 2.4 to 1 for the most severe fluency problems, and schools had flagged only 1 in 4 of the impaired boys and 1 in 7 of the impaired girls as learning disabled (Journal of Learning Disabilities, Quinn and Wagner, Oct 2013).
After matching on reading scores and background, English boys had 1.18 times the odds of girls of being identified with a specific learning difficulty, rising to 1.98 times when the model matched on maths scores (Journal of Learning Disabilities, Daniel et al., Feb 2026).
In Iran, 7.5% of primary school boys had dyslexia against 4% of girls (Iranian Journal of Psychiatry, Sakhai et al., Apr 2025).

National records point the same way, with a smaller gap than referral-based samples once suggested.

In the US in 2025, 10.3% of boys aged 3 to 17 had been told they have a learning disability, against 7.4% of girls (CDC National Center for Health Statistics, NHIS, Aug 2026).
The US gap was wider a decade earlier: across 2009 to 2017, 9.56% of boys and 5.85% of girls had a learning disability (Pediatrics, Zablotsky et al., Oct 2019).
In England, boys make up 53.0% of pupils whose primary need is a specific learning difficulty, a far smaller majority than the 60.4% of all SEN support pupils and 70.7% of EHC plan pupils who are male (Department for Education, Special educational needs in England, Jun 2026).
Scotland's count is close to even too: 20,521 boys and 18,515 girls had dyslexia recorded as a reason for support in 2025, so boys are 52.6% of the total (Scottish Government, Pupil census supplementary statistics, Dec 2025).

Is dyslexia genetic, and what conditions come with it?

Dyslexia runs strongly in families, and a large share of children with it also have ADHD or maths difficulties. A reading score on its own understates what a student with dyslexia is dealing with.

Manhattan plot from a genome-wide study of 51,800 adults with dyslexia and 1,087,070 controls, showing dozens of loci above the genome-wide significance line across chromosomes 1 to 22 and X — Figure 2: Each red peak above the grey line is a region of the genome tied to dyslexia, and none of them comes close to explaining it alone, which is what a highly polygenic trait looks like. Source: Nature Genetics, Doust et al., Discovery of 42 genome-wide significant loci associated with dyslexia, Oct 2022. Chart published by the source, captured 2 Oct 2026.

Is dyslexia hereditary?

Yes. Dyslexia clusters in families, and genetic studies find it spread across many genes of small effect rather than a single gene.

The largest dyslexia genome-wide study to date compared 51,800 adults who reported a dyslexia diagnosis with 1,087,070 controls and found 42 independent genome-wide significant loci, 27 of them new (Nature Genetics, Doust et al., Oct 2022).
Common genetic variants measured in that study explain about 19% of the liability to dyslexia, well below the 40% to 80% heritability that twin studies report, a gap the authors attribute partly to rare and structural variants (Nature Genetics, Doust et al., Oct 2022).
Dyslexia polygenic scores explained up to 6% of the variance in reading traits in independent cohorts (Nature Genetics, Doust et al., Oct 2022).
Family-risk rates varied widely by study, from 29% in a Dutch sample to 66% in an English one (Psychological Bulletin, Snowling & Melby-Lervag, Jan 2016).
A 2025 analysis that combined the 23andMe dyslexia study with GenLang reading data raised the effective sample to 1,228,832 people and identified 80 independent genome-wide significant loci, 36 not previously reported as significant (Translational Psychiatry, Aug 2025).
A polygenic index built from that analysis predicted 2.34% to 4.73% of the variance in reading traits among 6,410 members of the UK National Child Development Study cohort (Translational Psychiatry, Aug 2025).

Genome-wide and twin studies of reading skills point the same way.

In a study of up to 33,959 people aged 5 to 26 tested on word reading, nonword reading, spelling, phoneme awareness and nonword repetition, common variants accounted for 13% to 26% of the variation in each skill (PNAS, Eising et al., Aug 2022).
A meta-analysis of 49 twin studies covering more than 38,000 children aged 4 to 18 found genes outweighed environment for most reading skills, and that sex and spoken language did not change the heritability of any of them (Neuroscience & Biobehavioral Reviews, Andreola et al., Nov 2020).
In the Colorado Longitudinal Twin Study of Reading Disability, shared genes accounted for 86% of the stability of reading scores over a 5- to 6-year gap in twins with a history of reading difficulties (Annals of Dyslexia, Wadsworth et al., Dec 2007).
Genetic liability to dyslexia runs against genetic liability for educational attainment (genetic correlation of -0.186), and in 4,274 school-aged children the two polygenic scores pushed school performance in opposite directions (Translational Psychiatry, May 2026).

How often does dyslexia occur with ADHD or math difficulties?

Often. A sizeable share of children with dyslexia also have ADHD or maths difficulties, and the overlap runs well above what chance would predict. For the ADHD side of the overlap, see our ADHD statistics.

Adults with a dyslexia diagnosis were more than twice as likely to report ADHD as controls: 24% against 9% (Nature Genetics, Doust et al., Oct 2022).
A review of 17 studies published from 2001 to 2011 put the mean comorbidity rate between ADHD and learning disabilities at 45.1%, higher than earlier estimates, possibly because newer studies counted writing disorders (Journal of Learning Disabilities, DuPaul, Gormley & Laracy, Nov 2012).
In a Colorado sample, 32% of children with a reading disability alone met criteria for an ADHD subtype, against 7% of children with neither a reading nor a math disability (Journal of Learning Disabilities, Willcutt et al., Feb 2013).
Students with both a reading disability and ADHD were far more likely to have an IEP or special education services than students with a reading disability alone: 74% against 38% (Mind, Brain, and Education, Willcutt et al., Oct 2023).

Maths and spelling problems overlap with reading problems at three to five times the chance rate.

In a representative sample of 1,633 German-speaking third and fourth graders, learning disorders in reading, spelling and arithmetic co-occurred three to five times more often than chance would predict (PLoS ONE, Moll et al., Jul 2014).
In the same sample, 58% of children with a reading deficit also had a deficit in spelling or arithmetic at the 1 standard deviation cutoff, and 47% did at the stricter 1.5 standard deviation cutoff (PLoS ONE, Moll et al., Jul 2014).
In a population sample of 2,586 elementary school children, rates of arithmetic, reading or spelling deficits were four to five times higher among children already struggling in one of those areas (Journal of Child Psychology and Psychiatry, Landerl & Moll, Sep 2009).
In a screening of 691 Spanish fifth and sixth graders, 23.7% of children at risk for dyslexia also showed math difficulties, while 48.3% of children at risk for dyscalculia also had reading difficulties (Frontiers in Psychology, Oct 2025).

Is dyslexia linked to anxiety or depression?

Yes, as an association. Poor readers and students with learning disabilities score higher on anxiety and depression measures than their peers, and these studies describe risk, not cause.

A meta-analysis of 34 studies covering 16,275 participants (2,491 poor readers) found poor readers scored higher than typical readers on anxiety (effect size d = 0.41) and depression (d = 0.23) (Clinical Psychology Review, Francis et al., Sep 2018).
Across 58 studies, school-aged students with learning disabilities had higher anxiety scores than peers without them, a medium effect (d = 0.61) (Journal of Learning Disabilities, Nelson & Harwood, Apr 2010).
10% of children with a reading disability alone met criteria for major depression, against 2% of children with no reading or math disability; the rate reached 22% when reading and math disabilities occurred together (Journal of Learning Disabilities, Willcutt et al., Feb 2013).
14% of children with a reading disability alone met criteria for generalized anxiety disorder, against 6% of controls and 24% of children with both reading and math disabilities (Journal of Learning Disabilities, Willcutt et al., Feb 2013).
Among students with a reading disability, girls were about twice as likely as boys to meet criteria for major depressive disorder, 23% against 11% (Mind, Brain, and Education, Willcutt et al., Oct 2023).
In a community twin sample of 209 children with a reading disability and 192 without, the link between reading disability and symptoms of anxiety and depression held after controlling for ADHD, while links with aggression and conduct problems did not (Journal of Child Psychology and Psychiatry, Willcutt & Pennington, Nov 2000).
In a study following 208 Australian children (61 with dyslexia) across the move to secondary school, dyslexia had no direct effect on Year 7 depression or anxiety once earlier symptoms were accounted for; the link ran through weaker connection to school and peers (Dyslexia, Sumner et al., Jul 2026).

How do students with dyslexia do in reading and writing?

The reading slide in US schools is broad, and students with disabilities sit at the bottom of a distribution that was falling before the pandemic. Writing is where the gap is widest.

Share of US fourth-graders at each NAEP reading achievement level in 1992, 2019, 2022 and 2024, with the share below NAEP Basic growing since 2019 — Figure 3: The share of US fourth-graders below NAEP Basic in reading has grown in each assessment since 2019, while the share at NAEP Proficient or above has shrunk. Source: NCES, The Nation's Report Card, NAEP Reading 2024: National Trends, Jan 2025. Screenshot of the source page, captured 2 Oct 2026.

How many US students read below grade level?

NAEP does not measure grade level, but its lowest benchmark, NAEP Basic, is the closest marker, and a growing share of US students score below it. Students with disabilities, a group that includes many but not only students with dyslexia, are far more likely to score below it than their classmates.

40% of US fourth-graders performed below NAEP Basic in reading in 2024, a larger share than in 2022 (NCES, The Nation's Report Card, Jan 2025).
33% of US eighth-graders scored below NAEP Basic in reading in 2024, up from 27% in 2019 (NCES, The Nation's Report Card, Jan 2025).
32% of US twelfth-graders read below NAEP Basic in 2024, 12 percentage points more than in 1992 (NCES, The Nation's Report Card, Sep 2025).
66% of US eighth-graders with disabilities (an IEP or a Section 504 plan) scored below NAEP Basic in reading in 2024, against 28% of eighth-graders without a disability (NCES, The Nation's Report Card, Jan 2025).
Average reading scores for US 9-year-olds fell 5 points between 2020 and 2022, the largest drop in reading since 1990 (NCES, The Nation's Report Card, Sep 2022).
Reading scores for US 13-year-olds fell 4 points from 2020 to 2023 and 7 points over the decade (NCES, The Nation's Report Card, Jun 2023).

How do students with learning disabilities score in writing?

Worse than in any other subject. English pupils with a specific learning difficulty trail their classmates by more in writing than in reading, and US students with disabilities reach NAEP Proficient in writing at a fraction of the overall rate.

5% of US eighth-graders with disabilities performed at or above NAEP Proficient in writing in 2011, the latest national writing assessment, against 27% of all eighth-graders (NCES, The Nation's Report Card, Sep 2012).
45% of Year 6 pupils in England with a specific learning difficulty as their primary need met the expected reading standard in 2026, against 85% of pupils with no special educational need (Department for Education, Key stage 2 attainment, Sep 2026).
31% of Year 6 pupils in England with a specific learning difficulty met the expected standard in grammar, punctuation and spelling in 2026, against 85% of pupils with no special educational need (Department for Education, Key stage 2 attainment, Sep 2026).
23% of Year 6 pupils in England with a specific learning difficulty met the expected standard in reading, writing and maths combined in 2026, against 75% of pupils with no special educational need (Department for Education, Key stage 2 attainment, Sep 2026).
Of the Year 6 pupils in England working below the curriculum in at least one subject in 2026, 48% (25,196 pupils) were below it only in writing, against 1% only in reading and 1% only in maths (Department for Education, Key stage 2 attainment, Sep 2026).
At GCSE, 39.8% of the 19,518 pupils in England whose primary need is a specific learning difficulty achieved grade 4 or above in both English and maths in 2024/25, against 72.6% of pupils with no identified special educational need (Department for Education, Key stage 4 performance, Apr 2026).

The research on how dyslexic students write points at spelling and transcription, not ideas or vocabulary, as the bottleneck.

Spelling ability and within-word pausing together explained more than 76% of the variance in handwriting output among 9-year-olds with dyslexia, who formed letters as fast as their peers but wrote less and paused more (Sumner, Connelly & Barnett, Journal of Experimental Psychology: Learning, Memory, and Cognition, Sep 2014).
Spelling ability and pausing predicted 53% of the variance in the range of words children with dyslexia used when writing, while their spoken vocabulary had no effect (Sumner, Connelly & Barnett, Journal of Learning Disabilities, Sep 2014).
University students with dyslexia made more spelling errors and had their essays rated lower than 32 matched peers, even though they spent as long writing and produced as much text (Sumner & Connelly, Journal of Learning Disabilities, Jan 2020).
Children with dyslexia (average age 9.5) made more handwriting legibility errors than age-matched peers on both a dictation task and a simple alphabet task, and wrote the alphabet more slowly (Van Heuverswyn, Gosse & Van Reybroeck, Dyslexia, Apr 2024).

Does early reading intervention close the gap?

It narrows the gap but rarely closes it. Interventions for elementary students with or at risk for dyslexia produce small to moderate gains, effects are larger when instruction starts early, and the gap seen in first grade can still be measured at age 42.

In the National Reading Panel's meta-analysis, phonics instruction had about twice the effect when it began in kindergarten (d = 0.56) or first grade (d = 0.54) as when it began in grades 2 to 6 (d = 0.27) (NICHD, National Reading Panel, Apr 2000).
In England, 43% of Year 1 pupils with special educational needs met the expected standard in the 2025 phonics screening check, against 80% of all pupils (Department for Education, Phonics screening check attainment, Oct 2025).
In a Connecticut cohort followed from age 5 to age 42, first-grade reading level and reading growth through fifth grade explained 61% of the variance in adult reading accuracy among dyslexic readers (Ferrer, Shaywitz, Holahan & Shaywitz, npj Science of Learning, Nov 2023).
A meta-analysis of 53 studies covering 6,053 students in kindergarten to grade 5 with or at risk for dyslexia, run between 1980 and 2020, found an average effect of g = 0.33 on norm-referenced reading tests (Hall et al., Reading Research Quarterly, Sep 2022).
In the same meta-analysis, studies in kindergarten to grade 2 averaged g = 0.36 and studies in grades 3 to 5 averaged g = 0.16, though the difference was not statistically significant (Hall et al., Reading Research Quarterly, Sep 2022).
Interventions that taught spelling alongside word reading had larger effects (g = 0.37) than those without a spelling component (g = 0.23) (Hall et al., Reading Research Quarterly, Sep 2022).
Interventions described as multisensory (g = 0.20) did not produce significantly different effects from those that were not (g = 0.34), in an analysis the authors call underpowered (Hall et al., Reading Research Quarterly, Sep 2022).

Intensity and group size matter, and older struggling readers gain less on standardized tests.

Intensive reading interventions for struggling readers in kindergarten to grade 3 produced an average effect of 0.39 across 25 studies, or 0.28 after adjusting for publication bias (Wanzek et al., Journal of Learning Disabilities, May 2018).
One-on-one intensive intervention averaged an effect of 0.59, against 0.33 for groups of two to eight students, which the authors describe as limited evidence (Wanzek et al., Journal of Learning Disabilities, May 2018).
For struggling readers in grades 4 to 12, interventions studied from 1980 to 2011 averaged an effect of 0.49, falling to 0.21 on standardized tests (Scammacca et al., Journal of Learning Disabilities, Oct 2013).

What do learning disability statistics show about special education?

Specific learning disability is the largest category in US special education, but it is a legal eligibility bucket, not a diagnosis. States draw its edges differently, and for most students in it the label means support inside a general classroom, not a separate track.

Bar chart of the percentage distribution of disability types among US students ages 3 to 21 served under IDEA in school year 2022-23: specific learning disability 32%, speech or language impairment 19%, other health impairment 15%, autism 13%, developmental delay 7%, intellectual disability 6%, emotional disturbance 4%, multiple disabilities 2%, hearing impairment 1% — Figure 4: Specific learning disability outranks every other IDEA category, including speech or language impairment, other health impairment and autism, which makes it the single biggest driver of special education caseloads; it includes dyslexia, but also dyscalculia and other specific learning disabilities. Source: National Center for Education Statistics, Condition of Education 2024: Students With Disabilities, May 2024. Chart published by the source, captured 2 Oct 2026.

How many students receive special education for a specific learning disability?

About one in 20 public school students. Students with a specific learning disability made up 4.9% of total US public school enrollment in 2022-23, down from 6.1% in 2000-01 (NCES, Digest of Education Statistics, Table 204.30, Dec 2023).

The category is smaller than it was a generation ago: 2.86 million students, 45.4% of everyone served under IDEA, had a specific learning disability in 2000-01, against 2.41 million in 2022-23 (NCES, Digest of Education Statistics, Table 204.30, Dec 2023).
The share of the population aged 6 through 21 served under IDEA ranged from 6.4% to 15.1% across states in fall 2016, and because states set their own eligibility criteria, a child eligible in one state might be ineligible in another (U.S. Government Accountability Office, GAO-19-348, Apr 2019).
A review of learning disability rules in all 50 states and DC found that only 67% of states allowed the ability-achievement discrepancy method, 20% explicitly prohibited it, and about 16% required response to intervention as the sole identification model (School Psychology Quarterly, Maki, Floyd and Roberson, Jan 2015).

Who lands in the category also differs by sex and by race and ethnicity.

Specific learning disability is a bigger share of special education for girls than for boys: 41.2% of school-age girls served under IDEA were in the category in fall 2024, against 28.7% of boys, whose caseloads lean more toward autism and other health impairment (U.S. Department of Education, Office of Special Education Programs, IDEA Section 618 Child Count and Educational Environments 2024-25, Dec 2025).
Specific learning disability accounted for 38.6% of Hispanic and 34.5% of Black school-age students served under IDEA in fall 2024, against 30.1% of White and 17.1% of Asian students (U.S. Department of Education, Office of Special Education Programs, IDEA Section 618 Child Count and Educational Environments 2024-25, Dec 2025).

How much time do students with learning disabilities spend in general classrooms?

Most of the day. In fall 2024, 78.8% of school-age students served under IDEA for a specific learning disability spent 80% or more of the school day inside a regular class, and only 2.8% spent less than 40% of the day there (U.S. Department of Education, Office of Special Education Programs, IDEA Section 618 Child Count and Educational Environments 2024-25, Dec 2025).

Inclusion has risen for a decade: the share of all school-age students served under IDEA who spent 80% or more of their day in general classes rose from 61% in fall 2012 to 67% in fall 2022 (NCES, Condition of Education, May 2024).
In 1989, only 31.7% of school-age students served under IDEA spent 80% or more of their day in a general class (NCES, Digest of Education Statistics, Table 204.60, Oct 2023).
A separate group gets help without special education at all: 1,692,575 K-12 public school students, 3.5% of enrollment, were served solely under Section 504 of the Rehabilitation Act in 2021-22 (U.S. Department of Education, Office for Civil Rights, Civil Rights Data Collection 2021-22, Jan 2025).
Section 504 plans skew white: White students were 59.9% of students served only under Section 504 in 2021-22 but 45.2% of K-12 enrollment (U.S. Department of Education, Office for Civil Rights, Civil Rights Data Collection 2021-22, Jan 2025).
Scotland reports even higher mainstream inclusion: 274,367 of the 299,445 pupils with additional support needs in 2025, 92%, spent all of their time in mainstream classes (Scottish Government, Pupil census supplementary statistics, Dec 2025).

Do students with learning disabilities graduate from high school?

Most do, and more often than students with other disabilities, but on-time graduation for students with disabilities trails the national rate in every state.

Dropping out is still common: 25,102 students with a specific learning disability dropped out in 2023-24, 11.9% of those who left school, close to the 12.8% rate for all students with disabilities (U.S. Department of Education, Office of Special Education Programs, IDEA Section 618 Exiting 2023-24, Sep 2025).
On the four-year graduation measure, 71% of all students with disabilities graduated on time in 2021-22, against 87% of all US public high school students (NCES, Condition of Education, May 2024).
The on-time graduation rate for students with disabilities ranged from 49% in New Jersey to 84% in Florida in 2021-22, and was below the US average for all students in every state and DC (NCES, Condition of Education, May 2024).

In England, pupils with a specific learning difficulty are suspended at about twice the overall rate, mostly in secondary school.

7.0% of English pupils whose primary need is a specific learning difficulty were suspended at least once in 2024/25 (Department for Education, Suspensions and permanent exclusions in England, Jul 2026).
English secondary pupils with a specific learning difficulty had 32.1 suspensions per 100 pupils in 2024/25, against 3.6 for primary pupils with the same need (Department for Education, Suspensions and permanent exclusions in England, Jul 2026).
406 English pupils with a specific learning difficulty were permanently excluded in 2024/25, a rate of 0.23 per 100 pupils against 0.12 for all pupils (Department for Education, Suspensions and permanent exclusions in England, Jul 2026).
Most still move on to education or work: 89.5% of English pupils with a specific learning difficulty who finished key stage 4 in 2022/23 were in a sustained education, apprenticeship or employment destination the following year, against 92.9% of pupils with no identified SEN (Department for Education, Key stage 4 destination measures, Apr 2026).

How does dyslexia affect adults in college and at work?

Adults with low literacy or a learning difficulty have fewer qualifications and lower employment, and the workplace adjustments they need usually cost little.

Stacked bar charts of US adults ages 16 to 65 by PIAAC literacy and numeracy proficiency level in 2012/14, 2017 and 2023, with the share at Level 1 or below growing in 2023 — Figure 5: The lowest literacy tier barely moved between 2012/14 and 2017, then jumped by 2023, and numeracy slid with it, so more working-age adults now struggle with the written material most jobs depend on. The chart covers all US adults, not adults with dyslexia. Source: U.S. Department of Education, National Center for Education Statistics (NCES), Highlights of the 2023 U.S. PIAAC Results, Dec 2024. Screenshot of the source page, captured 2 Oct 2026.

How many college students have a learning disability?

In England, about one in 17: in 2023-24, 103,760 undergraduates at English higher education providers (5.9%) reported a cognitive or learning difficulty, the group that includes dyslexia, up from 70,570 (4.6%) in 2010-11 (Office for Students, Student characteristics population data, Oct 2025). The US counts all disabilities together, and most college students with a learning disability never tell their school, so official counts sit below the real number.

Counting every type of disability, including difficulty hearing, seeing, concentrating, remembering or walking, 20.5% of US undergraduates, about 3.5 million students, reported a disability in 2019-20 (NCES, Digest of Education Statistics Table 311.10, Jun 2023).
Within eight years of leaving high school, 67% of US young adults with learning disabilities had enrolled in some postsecondary school, but only 21% had ever attended a 4-year college (IES, National Longitudinal Transition Study-2, Sep 2011).
Among postsecondary students with learning disabilities, 69% said they did not consider themselves to have a disability (IES, National Longitudinal Transition Study-2, Sep 2011).
UK-domiciled full-time first-degree students in England with a cognitive or learning difficulty continued into their second year at a higher rate than students with no reported disability: 90.8% against 88.3% for 2022-23 entrants (Office for Students, Student characteristics outcomes data, Oct 2025).
After graduating, 73.9% of UK-domiciled full-time first-degree qualifiers with a cognitive or learning difficulty reported a positive outcome, such as professional work or further study, against 71.4% of those with no reported disability (2022-23 qualifiers) (Office for Students, Student characteristics outcomes data, Oct 2025).
In 2025, about 24% of Americans aged 25 and over with a disability had a bachelor's degree or higher, against about 42% of those with no disability (U.S. Bureau of Labor Statistics, Persons with a Disability: Labor Force Characteristics, Mar 2026).

How do adults with low literacy or a disability fare at work?

Both groups have lost ground. The share of US adults with low literacy and numeracy grew between 2017 and 2023, and people with disabilities are far less likely to be in work than everyone else in both the US and the UK.

Over 2017 to 2023, the share of US adults at Level 1 or below in numeracy rose from 29% to 34% (NCES, PIAAC 2023 U.S. National Results, Dec 2024).
Even among US adults with more than a high school education, the share at Level 1 or below in literacy doubled from 6% in 2017 to 13% in 2023 (NCES, PIAAC 2023 U.S. National Results, Dec 2024).
In 2025, 22.8% of Americans with a disability were employed, against 65.2% of those without a disability (U.S. Bureau of Labor Statistics, Persons with a Disability: Labor Force Characteristics, Mar 2026).
The US unemployment rate for people with a disability rose 0.8 points to 8.3% in 2025, double the 4.1% rate for people without one (U.S. Bureau of Labor Statistics, Persons with a Disability: Labor Force Characteristics, Mar 2026).
Workers with a disability were more likely to be self-employed than workers without one in 2025: 9.1% against 5.9% (U.S. Bureau of Labor Statistics, Persons with a Disability: Labor Force Characteristics, Mar 2026).
Up to eight years after high school, 67% of US young adults with learning disabilities were employed at the time of interview, against 50% of those with emotional disturbances (IES, National Longitudinal Transition Study-2, Sep 2011).

The UK data also shows lower qualifications among disabled adults with learning difficulties, and high rates of learning difficulty among prisoners.

The UK disability employment rate was 52.8% in April to June 2025, against 82.5% for non-disabled people, a gap of 29.7 percentage points (Department for Work and Pensions, The employment of disabled people 2025, Nov 2025).
About one in three (34.4%) UK disabled people with severe or specific learning difficulties had no qualifications in 2024/25, against 12.0% of all disabled people (Department for Work and Pensions, The employment of disabled people 2025, Nov 2025).
Where known, 57% of prisoners in England's public prisons who took an initial assessment and enrolled on a course had a learning difficulty and/or disability in the year to March 2025 (Ministry of Justice, Prison Education and Accredited Programme Statistics 2024 to 2025, Sep 2025).
65% of English initial assessments taken by prisoners in England's public prisons came out at Entry level 1 to 3, the bands below Level 1 (Ministry of Justice, Prison Education and Accredited Programme Statistics 2024 to 2025, Sep 2025).

What workplace accommodations do employees with dyslexia use?

Most adjustments are cheap or free. The bigger barrier is disclosure: most workers with a learning disability never tell their employer, so they never get the adjustment.

Only 6% of employers reported ongoing accommodation costs, with a median of $2,400 a year (Job Accommodation Network, Costs and Benefits of Accommodation, Sep 2025).
66% of employers rated the accommodations they made as very or extremely effective (Job Accommodation Network, Costs and Benefits of Accommodation, Sep 2025).
85% of employers that made an accommodation said it helped them keep the employee (Job Accommodation Network, Costs and Benefits of Accommodation, Sep 2025).
Among employed US young adults with learning disabilities, 19% said their employer knew about their disability and 5% had received an accommodation at work (IES, National Longitudinal Transition Study-2, Sep 2011).
31% of UK employees who identify as neurodivergent had told neither their line manager nor HR, in a survey of 1,047 employed adults of whom 31% identified as dyslexic (CIPD, Neuroinclusion at work report 2024, Feb 2024).
In a survey of 1,003 UK senior managers, only 24% said their organisation provides training or guidance for line managers on how to support neurodivergent team members (CIPD, Neuroinclusion at work report 2024, Feb 2024).

The UK's Access to Work scheme is one of the few programmes that counts dyslexia as its own condition.

5,550 people with dyslexia as their primary condition received an Access to Work payment in 2024/25, out of 74,190 recipients across all conditions (Department for Work and Pensions, Access to Work statistics April 2007 to March 2025, Jan 2026).
The average Access to Work payment for a customer with dyslexia was £2,700 in 2024/25, against £4,000 across all customers (Department for Work and Pensions, Access to Work statistics April 2007 to March 2025, Jan 2026).

Does assistive technology for dyslexia improve reading and writing?

Text-to-speech has the most research behind it among assistive tools for dyslexia, and it points to a small to moderate effect. Speech-to-text has less research behind it, much of it small or old. Both are access tools; neither teaches reading.

Forest plot of 23 effect sizes from 22 studies measuring the effect of text-to-speech and read-aloud tools on reading comprehension for students with reading disabilities, with study effects ranging from below zero to above 1 and a pooled random-effects estimate — Figure 6: Read-aloud and text-to-speech tools helped comprehension in most of the 22 studies (23 effect sizes), but results ran from clear losses to large gains, which is why the pooled effect is small to moderate rather than large. Source: Journal of Learning Disabilities, Wood et al., Does Use of Text-to-Speech and Related Read-Aloud Tools Improve Reading Comprehension for Students with Reading Disabilities? A Meta-Analysis, Jan 2017. Chart published by the source, captured 2 Oct 2026.

Does text-to-speech improve reading comprehension?

Yes, on average, by a small to moderate amount. Listening while reading raises comprehension scores for students with reading disabilities, but the gain shrinks in stricter study designs and varies widely from study to study.

After adjusting for four studies the authors judged likely missing through publication bias, the Wood meta-analysis put the effect at d = .24 (95% CI .02 to .45) (Journal of Learning Disabilities, Wood et al., Jan 2017).
Study design changed the answer: studies comparing separate groups found d = .61, while studies testing the same students with and without read-aloud found d = .15, an effect not statistically different from zero (Journal of Learning Disabilities, Wood et al., Jan 2017).
A meta-analysis of 114 effect sizes from 23 studies found read-aloud test accommodations helped students with disabilities significantly more than students without disabilities, and helped more when a human proctor read than when a computer or recording did (Educational Measurement: Issues and Practice, Li, Mar 2014).
For adolescents and adults with learning disabilities, a meta-analysis of text-to-speech studies found an overall effect on reading comprehension of g = 0.445 (p = 0.06), falling to 0.167 once one large outlier study was removed (Computers & Education, Perelmutter et al., Jun 2017).
In a study of children aged 8 to 12 with reading and language difficulties, both text-to-speech conditions produced significantly higher comprehension scores than silent reading, and children with dyslexia only gained more than children who also had a language impairment (Annals of Dyslexia, Keelor et al., Apr 2023).

Smaller trials show where the gains come from and where they stop.

Among 164 ninth-grade struggling readers in Hawaii, those who read course materials with text-to-speech software for 10 weeks (582 minutes on average) improved significantly on unassisted reading vocabulary and comprehension tests (Assistive Technology, Park et al., May 2016).
In a single-case study of four middle schoolers with reading learning disabilities, continuous text-to-speech gave three of them their best comprehension accuracy and rate, with effect sizes reaching 0.70 and 0.99 against silent reading (Assistive Technology, Schmitt et al., Mar 2018).
When an examiner read only the directions, names and questions aloud, the 28 poor decoders among 82 third graders scored significantly higher on a reading comprehension test, while the 54 average decoders did not benefit (Journal of Learning Disabilities, Giusto & Ehri, Jul 2018).
Forty-three Greek students with dyslexia identified words and sentences better in natural speech than in synthetic speech, but their comprehension of whole texts did not differ between the two (Disability and Rehabilitation: Assistive Technology, Giannouli & Banou, Jul 2019).
In a Swedish trial of 149 students with severe reading disabilities, a group given 24 sessions of assistive technology training improved as much as the normed population over a year, but no more than a control group receiving usual support (Disability and Rehabilitation: Assistive Technology, Svensson et al., Aug 2019).

Does speech-to-text help students with dyslexia write?

It helps most with getting words onto the page. Studies since 1995 find students with learning disabilities write longer texts with fewer spelling errors when they dictate, but gains in overall quality are less consistent, and the research base is small and partly decades old.

University students with learning disabilities (n = 29) earned significantly higher holistic essay scores when they dictated to a speech recognition system than when they wrote without assistance (Learning Disability Quarterly, Higgins & Raskind, May 1995).
High school students with and without learning disabilities learned to dictate with speech recognition at under 10% total word errors, and for the students with learning disabilities, dictated essays were better than handwritten ones (Exceptional Children, MacArthur & Cavalier, Oct 2004).
Among less fluent writers aged 11 to 14, speech recognition significantly lengthened narratives and cut surface errors compared with handwriting, but did not significantly improve rated quality (Journal of Educational Psychology, Quinlan, Jun 2004).
A systematic review of assistive technology for adolescents and adults with learning disabilities found speech-to-text effects ranging from 0.42 to 1.125 across studies, but the six studies (all using Dragon software) were too different to pool (Computers & Education, Perelmutter et al., Jun 2017).
A scoping review of speech-to-text research with secondary pupils aged 12 to 18 with learning difficulties found only eight peer-reviewed studies and five grey-literature reports from 2000 to April 2022, two of them experimental (Disability and Rehabilitation: Assistive Technology, Matre & Cameron, Nov 2022).
In a Nordic single-case study, seven of eight middle school students with writing difficulties produced more text with speech-to-text than with their keyboarding baseline, and word-level accuracy held steady or improved (Disability and Rehabilitation: Assistive Technology, Almgren Bäck et al., May 2024).
In the first UK group study of speech-to-text for children with special educational needs, 30 children with Education, Health and Care Plans used Dragon for 16 to 18 weeks, and their screen-written text at the end was significantly better than their handwriting (Research in Developmental Disabilities, Kambouri et al., Feb 2023).

The newest work looks at what goes wrong as well as what improves.

When four students aged 10 to 12 with severe dyslexia wrote with speech-to-text, their texts contained no conventional spelling errors, but misrecognised words and dropped words or sentence parts became the most frequent errors (Dyslexia, Svendsen et al., Feb 2026).
A review of nine widely used reading and writing programs for three Nordic languages found that the best spelling aids now let students with moderate spelling difficulties spell accurately, but those features take time and attention that can disrupt writing fluency (Dyslexia, Mossige et al., Sep 2023).

These studies measure the text students produce, not their underlying reading or spelling skill. For how dictation fits into homework, notes and essays, see voice to text for students.

How many students with dyslexia use assistive technology?

No national survey counts it. The closest measures are test accommodations and school surveys: most US students with disabilities take the national reading test with an accommodation, and English schools report that text-to-speech and dictation are common but hard to put into practice.

On the 2024 NAEP reading assessment, 63% of fourth graders and 70% of eighth graders identified as students with disabilities were assessed with accommodations (NCES, The Nation's Report Card, Jan 2025).
Among staff at English schools on a Department for Education assistive technology course, the share whose school used text-to-speech tools rose from 44% before the course to 83% after, and dictation tools from 38% to 72% (36 post-course respondents) (Department for Education, AT training for schools evaluation, Sep 2024).
Before that training, 91% of participants disagreed that their school used assistive technology to its full potential; afterwards 44% still disagreed (Department for Education, AT training for schools evaluation, Sep 2024).
In a December 2025 survey, 44% of primary and 42% of secondary school leaders in England said accessing or implementing assistive technology was a challenge in helping teachers deliver inclusive teaching (Department for Education, School and college voice, Apr 2026).
After a six-week reading and writing app intervention in Sweden, special education teachers reported that 82% of the younger and 47% of the older students with reading difficulties (59 students from Grade 4, Grade 8 and high school) were still using the technology, to varying degrees (Disability and Rehabilitation: Assistive Technology, Nordström et al., Sep 2018).
In five-year follow-up interviews with nine Swedish students with dyslexia, audiobooks were described as helpful throughout school, while speech-to-text was of mixed use (Disability and Rehabilitation: Assistive Technology, Almgren Bäck et al., Jan 2023).

UK university students get assistive technology through the Disabled Students' Allowance. The latest published processing data, for 2019/20, showed long waits.

The Student Loans Company's framework for Disabled Students' Allowance needs assessments, assistive technology and assistive technology training is worth up to about £400m over three years (Student Loans Company, Jul 2023).
In the 2019/20 processing year, the average Disabled Students' Allowance application took 115 days from application to final award, across 36,934 students (Student Loans Company, Equality Impact Assessment, Mar 2021).
A systematic review of AI-based interventions for students with learning disabilities published from 2022 to 2025 found 11 studies with 3,033 participants, six of them on dyslexia; all reported positive outcomes, but none was rated at low risk of bias (Brain Sciences, Jul 2025).

Glossary

ADHD: attention-deficit/hyperactivity disorder.
ALN: additional learning needs, the system Wales uses in place of SEN.
AT: assistive technology, tools such as text-to-speech, dictation and spelling aids.
EHC plan: education, health and care plan, a statutory support plan in England.
GCSE: General Certificate of Secondary Education, the exams English pupils take at 16.
IDEA: Individuals with Disabilities Education Act, the US special education law.
IEP: Individualized Education Program, the plan for a US student served under IDEA.
IES: Institute of Education Sciences, the research arm of the U.S. Department of Education.
JAN: Job Accommodation Network, funded by the U.S. Department of Labor.
NAEP: National Assessment of Educational Progress, also called The Nation's Report Card.
NCES: National Center for Education Statistics.
NHIS: National Health Interview Survey, run by the CDC's National Center for Health Statistics.
PIAAC: Programme for the International Assessment of Adult Competencies, the adult skills survey.
SEN: special educational needs (England).
504 plan: support under Section 504 of the Rehabilitation Act for a US student who does not need special education.

Methodology and Sources

Every statistic on this page was traced to the organisation that produced the data: government agencies and regulators (NCES, the U.S. Department of Education, CDC, BLS, GAO, England's Department for Education, the Scottish and Welsh Governments, DWP, the Ministry of Justice, the Office for Students), peer-reviewed journals, and research bodies when they ran the study themselves (JAN, CIPD). We do not cite statistics roundups or advocacy pages that repeat figures without a study behind them. Each link was opened on 2 October 2026 to confirm the page loads, the number appears on it and the date is right; where a number lives in a data file, it was read from the file the page publishes. Percentages from the IDEA Section 618 child count and exiting files are our calculations from the counts in those files, which cover the US, outlying areas and freely associated states. Journal articles are dated by their first publication, usually the online date. Each figure says what its source measured, because learning disability counts, special education categories and dyslexia diagnoses are different things. Older studies are kept where they remain the best evidence and are dated so readers can judge them. The page is refreshed monthly, with new survey and school data swapped in as they are released.

167 Dyslexia Statistics October 2026