In its 2025 report, the CDC’s Autism and Developmental Disabilities Monitoring (ADDM) Network found that 1 in 31 US 8-year-olds (3.2%) was identified with autism in 2022 - up from 1 in 36 two years earlier and 4.8 times the rate the network recorded in 2000. Worldwide, the Global Burden of Disease Study 2021 counted 61.8 million autistic people, roughly 1 in 127 persons (The Lancet Psychiatry, 2024). About 5.4 million US adults, or 2.21%, are estimated to be on the spectrum (Dietz et al., CDC). And an estimated 25% to 30% of autistic children are minimally verbal or nonspeaking (multiple studies). This analysis consolidates data from the CDC ADDM Network, the World Health Organization, the Global Burden of Disease Study 2021, the NIH’s NIMH, and 15 other primary sources to show how prevalence is measured, why the numbers keep rising, and what they mean for diagnosis, communication, and adult life.
TL;DR
- 1 in 31 US 8-year-olds (3.2%) was identified with autism in 2022 (CDC ADDM Network, 2025).
- That is 4.8 times the 1 in 150 rate recorded in 2000 (NIMH; CDC ADDM Network).
- The CDC attributes most of the rise to better identification, broader criteria, and expanded screening, not a true epidemic (CDC ADDM Network, 2025).
- Prevalence ranged from 1 in 103 in Texas (Laredo) to 1 in 19 in California across 16 sites (CDC ADDM Network, 2025).
- Autism was 3.4 times more prevalent among boys than girls (49.2 vs 14.3 per 1,000) (CDC ADDM Network, 2025).
- 39.6% of identified 8-year-olds had a co-occurring intellectual disability (CDC ADDM Network, 2025).
- The median age of earliest diagnosis was 47 months (CDC ADDM Network, 2025).
- An estimated 61.8 million people, about 1 in 127, were autistic worldwide in 2021 (GBD Study 2021, The Lancet Psychiatry).
- Roughly 25-30% of autistic children are minimally verbal or nonspeaking (multiple studies).
- About 5.4 million US adults (2.21%) are on the autism spectrum (Dietz et al., CDC).
- Lifetime support costs reach $2.4 million with intellectual disability and $1.4 million without (Buescher et al., 2014).
- In England, 212,964 people had an open referral for suspected autism in December 2024, 90% waiting beyond the 13-week guideline (NHS England Digital, 2025).
1. US autism prevalence now
Autism is common and, on current surveillance, still rising. The CDC ADDM Network identified 1 in 31 US 8-year-olds (3.2%, or 32.2 per 1,000) with autism in 2022 - the highest rate in the network’s history. The figure is a surveillance count from 16 sites, not a national census, and it varies enormously by place: from about 1 in 103 in one Texas site to 1 in 19 in California. Boys are identified far more often than girls, though that gap is under active revision (see Section 3). The younger cohort tells the more forward-looking story - 4-year-olds are already being identified at 1 in 34.
| Metric | Value | Source |
|---|---|---|
| US 8-year-olds identified with autism (2022) | 1 in 31 (3.2%) | CDC ADDM Network, MMWR SS-7402, 2025 |
| Prevalence per 1,000 (age 8) | 32.2 | CDC ADDM Network, 2025 |
| US 4-year-olds identified with autism (2022) | 1 in 34 (2.9%) | CDC ADDM Network, 2025 |
| Prevalence, boys vs girls (per 1,000) | 49.2 vs 14.3 | CDC ADDM Network, 2025 |
| Male-to-female ratio (identified) | 3.4 to 1 | CDC ADDM Network, 2025 |
| Site range (age 8) | 9.7 (TX, Laredo) to 53.1 (California) per 1,000 | CDC ADDM Network, 2025 |
| US children 3-17 with an ASD diagnosis (parent-reported) | 3.2% | NSCH 2020-2021 (Li et al., 2025) |
Context: the near six-fold spread across sites reflects differences in screening, services, and record access, not a real difference in who is autistic. The ADDM Network is the canonical US anchor because it reviews health and school records rather than relying on parent report alone.
2. The trend, and why the numbers keep rising
The long climb is real, but its cause is mostly measurement. US prevalence has risen from 1 in 150 in 2000 to 1 in 31 in 2022 - a 4.8-fold increase that the CDC attributes primarily to broader diagnostic criteria, greater awareness, and expanded screening and services. The clearest evidence that identification (not occurrence) is driving the trend: children born in 2018 were identified with autism by age 4 at a markedly higher rate than those born in 2014, with early cumulative incidence 1.7 times as high. This is a story about a system getting better at seeing autistic children, and it argues against epidemic framing.
| Metric | Value | Source |
|---|---|---|
| Prevalence, 2000 surveillance year | 1 in 150 | NIMH (ADDM series) |
| Prevalence, 2010 | 1 in 68 | NIMH (ADDM series) |
| Prevalence, 2018 | 1 in 44 | NIMH (ADDM series) |
| Prevalence, 2020 | 1 in 36 | NIMH (ADDM series) |
| Prevalence, 2022 | 1 in 31 (32.2 per 1,000) | CDC ADDM Network, 2025 |
| 2022 rate vs first ADDM survey | 4.8x higher | CDC ADDM Network, 2025 |
| Early identification by age 4, 2018 vs 2014 birth cohorts | 1.7x higher cumulative incidence | CDC ADDM Network, 2025 |
Historical note: estimates from the early 1990s put autism nearer 1 in 1,500 before diagnostic criteria widened (NIMH). The direction of travel reflects recognition catching up with a population that was always there. For a parallel neurodivergence trend where identification has similarly reshaped the numbers, see our dyslexia statistics.
3. Demographics: sex, race, and intellectual disability
The 2022 data mark a demographic turning point. For the first time in ADDM history, autism prevalence was lower among non-Hispanic White children (27.7 per 1,000) than among Asian or Pacific Islander (38.2), American Indian or Alaska Native (37.5), Black (36.6), and Hispanic (33.0) children - a reversal of the historical pattern. Researchers read this as improved access and identification in previously underserved groups, not a change in underlying risk. Intellectual disability co-occurs in about two in five identified children, and the sex ratio, long quoted as 4 to 1, is now estimated closer to 3 to 1 once girls are actively assessed.
| Metric | Value | Source |
|---|---|---|
| Prevalence, Asian/Pacific Islander children (per 1,000) | 38.2 | CDC ADDM Network, MMWR SS-7402, 2025 |
| Prevalence, Black children (per 1,000) | 36.6 | CDC ADDM Network, 2025 |
| Prevalence, non-Hispanic White children (per 1,000) | 27.7 | CDC ADDM Network, 2025 |
| Co-occurring intellectual disability (overall) | 39.6% | CDC ADDM Network, 2025 |
| Co-occurring intellectual disability, Black children | 52.8% | CDC ADDM Network, 2025 |
| True male-to-female ratio (active assessment) | ~3 to 1 | Loomes et al., 2017 meta-analysis |
| Additional girls potentially undiagnosed (modeling) | up to 39% | Frontiers in Psychiatry, 2025 |
Outlier: intellectual disability co-occurrence ranged from 31.2% in multiracial children to 52.8% in Black children (CDC ADDM Network, 2025), a spread that tracks with disparities in who reaches diagnosis early versus late. Girls are also diagnosed later than boys despite comparable parental concern (Frontiers in Psychiatry, 2025).
4. Diagnosis and early identification
Autism can be reliably identified in toddlers, but most children are diagnosed years later. The median age of earliest known autism diagnosis was 47 months - just under 4 years - despite reliable signs appearing by 18 to 24 months. The lag is uneven: earliest-diagnosis medians ranged from 36 months in California to 69.5 months in one Texas site. Capacity is a core bottleneck. In a US survey of autism diagnostic centers, most reported waits of over four months from referral to evaluation, and a fifth had waitlists exceeding a year or were closed to new referrals.
| Metric | Value | Source |
|---|---|---|
| Median age of earliest diagnosis | 47 months | CDC ADDM Network, MMWR SS-7402, 2025 |
| Range of earliest-diagnosis median across sites | 36 to 69.5 months | CDC ADDM Network, 2025 |
| Children evaluated by age 3 (age-8 cohort) | ~55% | CDC ADDM Network, 2025 |
| Autism centers reporting waits over 4 months | nearly two-thirds | iSPI / Cognoa survey of US autism centers |
| Centers with waitlists over a year or closed | 21% | iSPI / Cognoa survey |
| Age at which reliable signs can appear | 18-24 months | CDC ADDM Network |
Why it matters: earlier identification opens the window for support at the ages when it is most effective. The same diagnostic capacity that gates autism assessment also gates access to communication and developmental services - see our speech therapy statistics for the workforce side of that pipeline.
5. Communication and co-occurring conditions
Autism is a spectrum, and communication profiles span the full range. An estimated 25% to 30% of autistic children are minimally verbal or nonspeaking, meaning they use few functional words or cannot rely on speech alone (multiple studies). Minimally verbal is not the same as non-thinking, and it is not always permanent - many children gain spoken language later, and many communicate fluently through augmentative and alternative communication (AAC), typing, or speech-generating devices. Co-occurring conditions are the rule rather than the exception, which is why autism care is often about the accompanying conditions as much as autism itself.
| Metric | Value | Source |
|---|---|---|
| Autistic children who are minimally verbal or nonspeaking | 25-30% | Socioeconomic factors in minimally verbal ASD (PMC) |
| Autistic children who experience language regression | ~20-25% | Pickles et al., JCPP 2022 |
| Median age of language regression | ~18 months | Pickles et al., JCPP 2022 |
| Epilepsy in autistic people | 25-40% (vs 2-3% general) | Children’s Hospital of Philadelphia |
| Children with autism who have a co-occurring psychiatric diagnosis | up to 85% | Children’s Hospital of Philadelphia |
| Range of intellectual functioning | profound impairment to superior | WHO, 2023 |
Context: estimates vary because “minimally verbal” has no single clinical threshold - studies have labeled children ranging from zero spoken words to a few dozen. For nonspeaking autistic people, communication access often runs through AAC; our AAC device statistics cover that market and the roughly 30-50% device-abandonment problem that shapes outcomes.
6. Autism around the world
Global numbers are larger in absolute terms and far noisier in quality. The Global Burden of Disease Study 2021 estimated 61.8 million autistic people worldwide - about 1 in 127 persons - and ranked autism among the top-ten causes of non-fatal health burden for people under 20. The WHO uses the same 1 in 127 figure for 2021 while cautioning that prevalence in many low- and middle-income countries is simply unknown. Systematic reviews find estimates spanning two orders of magnitude, driven by differences in criteria, awareness, and diagnostic infrastructure rather than true variation in occurrence.
| Metric | Value | Source |
|---|---|---|
| People autistic worldwide (2021) | 61.8 million (~1 in 127) | GBD Study 2021, The Lancet Psychiatry |
| Age-standardized global prevalence | 788.3 per 100,000 | GBD Study 2021 |
| Global prevalence, males vs females (per 100,000) | 1,064.7 vs 508.1 | GBD Study 2021 |
| Global disability burden | 11.5 million DALYs | GBD Study 2021 |
| WHO reference prevalence (2021) | 1 in 127 persons | World Health Organization, 2023 |
| Median estimate across studies | 100 per 10,000 (range 1.09-436.0) | Zeidan et al., Autism Research 2022 |
Flag: the Zeidan review’s 100-per-10,000 median (about 1%) is the most recent global synthesis, but its range - from roughly 1 to 436 per 10,000 - is a measure of how much diagnostic access, not biology, shapes reported rates. Higher-prevalence countries are generally those identifying more of their autistic population, not producing more autistic people.
7. Autistic adults: identification, work, and cost
Most autism data stop at childhood, which leaves autistic adults comparatively uncounted. A CDC study led by Dietz and colleagues estimated 5.4 million US adults (2.21%) are on the autism spectrum, yet employment outcomes remain poor - national indicators suggest only about 14% to 16% of autistic adults work full time. The gap is widely attributed to barriers, disclosure risk, and missing accommodations rather than capability; autistic adults who disclosed their diagnosis to an employer were reported to be several times more likely to be employed. The lifetime cost figures below are dominated by support and lost productivity, both of which early, well-matched services can reduce.
| Metric | Value | Source |
|---|---|---|
| US adults on the autism spectrum | 5.4 million (2.21%) | Dietz et al., CDC (National and State Estimates) |
| State range for adult prevalence | 1.97% (LA) to 2.42% (MA) | Dietz et al., CDC |
| Autistic adults in full-time paid work | ~14-16% | National Autism Indicators Report (Drexel) |
| Lifetime cost, autism with intellectual disability | $2.4 million | Buescher et al., JAMA Pediatrics 2014 |
| Lifetime cost, autism without intellectual disability | $1.4 million | Buescher et al., 2014 |
| England open referrals for suspected autism (Dec 2024) | 212,964 | NHS England Digital, 2025 |
| England referrals waiting beyond 13-week guideline | 90.0% (191,656) | NHS England Digital, 2025 |
Context: England’s autism waiting list rose 23% in a single year and 82% over two years (NHS England Digital, 2025), a demand surge mirrored in adult self-referral across many high-income systems. Community research also shows a clear language preference among autistic adults - a majority endorse identity-first terms such as “autistic person,” with the word “autism” endorsed by 91-96% across countries (Flowers et al., 2023). For how education access shapes the adult pipeline, see our online learning statistics; for the assistive-technology access parallel, see our screen reader statistics.
Summary: Autism Prevalence by the Numbers
| Metric | Value | Source |
|---|---|---|
| US 8-year-olds identified with autism (2022) | 1 in 31 (3.2%) | CDC ADDM Network, 2025 |
| Prevalence per 1,000 (age 8) | 32.2 | CDC ADDM Network, 2025 |
| US 4-year-olds identified with autism (2022) | 1 in 34 | CDC ADDM Network, 2025 |
| 2022 rate vs first ADDM survey (2000) | 4.8x higher (from 1 in 150) | NIMH; CDC ADDM Network |
| Male-to-female ratio (identified) | 3.4 to 1 | CDC ADDM Network, 2025 |
| Site range (age 8) | 1 in 103 to 1 in 19 | CDC ADDM Network, 2025 |
| Prevalence, White vs Asian/Pacific Islander (per 1,000) | 27.7 vs 38.2 | CDC ADDM Network, 2025 |
| Co-occurring intellectual disability | 39.6% | CDC ADDM Network, 2025 |
| Median age of earliest diagnosis | 47 months | CDC ADDM Network, 2025 |
| Children evaluated by age 3 | ~55% | CDC ADDM Network, 2025 |
| Autistic children minimally verbal or nonspeaking | 25-30% | Multiple studies |
| Epilepsy in autistic people | 25-40% (vs 2-3%) | Children’s Hospital of Philadelphia |
| People autistic worldwide (2021) | 61.8 million (~1 in 127) | GBD Study 2021, The Lancet Psychiatry |
| Global disability burden | 11.5 million DALYs | GBD Study 2021 |
| WHO reference prevalence (2021) | 1 in 127 | WHO, 2023 |
| Global median across studies | 100 per 10,000 (~1%) | Zeidan et al., 2022 |
| US adults on the spectrum | 5.4 million (2.21%) | Dietz et al., CDC |
| Autistic adults in full-time work | ~14-16% | National Autism Indicators Report |
| Lifetime cost with intellectual disability | $2.4 million | Buescher et al., 2014 |
| England referrals waiting beyond 13 weeks | 90.0% | NHS England Digital, 2025 |
Methodology and Sources
Data was gathered by aggregating figures from government surveillance reports, peer-reviewed studies, and international health datasets published or current as of July 2026, with US prevalence anchored to the CDC ADDM Network, global figures cross-referenced across the WHO and the Global Burden of Disease Study, and older clinical figures flagged where they are the most recent available.
Sources cited:
- CDC Autism and Developmental Disabilities Monitoring (ADDM) Network, MMWR SS-7402 (2025), surveillance year 2022
- CDC ADDM Network full report (PMC)
- National Institute of Mental Health (NIMH), Autism Spectrum Disorder statistics
- World Health Organization, Autism fact sheet (2023, referencing 2021 prevalence)
- Global Burden of Disease Study 2021, The Lancet Psychiatry (2024), global epidemiology of the autism spectrum
- Zeidan, J. et al. (2022), Global prevalence of autism: a systematic review update, Autism Research
- Loomes, R. et al. (2017), What is the male-to-female ratio in autism spectrum disorder? A systematic review and meta-analysis
- Frontiers in Psychiatry (2025), Female gender and autism: underdiagnosis and misdiagnosis
- iSPI / Cognoa, survey of wait times and processes for autism diagnostic evaluations at US autism centers
- Socioeconomic factors associated with minimally verbal status in individuals with ASD (PMC)
- Pickles, A. et al. (2022), Predictors of language regression and its association with subsequent communication development, Journal of Child Psychology and Psychiatry
- Children’s Hospital of Philadelphia, Autism’s Clinical Companions: Frequent Comorbidities with ASD
- Dietz, P.M. et al., National and State Estimates of Adults with Autism Spectrum Disorder (CDC)
- National Survey of Children’s Health (NSCH), 2020-2021 analysis of ASD prevalence in children aged 3-17
- Buescher, A.V.S. et al. (2014), Costs of Autism Spectrum Disorders in the United Kingdom and the United States, JAMA Pediatrics (CDC-hosted forecast for annual burden)
- National Autism Indicators Report (A.J. Drexel Autism Institute), employment outcomes
- NHS England Digital, Autism Statistics, January 2024 to December 2024
- Flowers, J. et al. (2023), Words matter: language preferences in a sample of autistic adults, Autism & Developmental Language Impairments
Data watch: the CDC ADDM Network publishes autism prevalence on a two-year cycle, so a next release covering 2024 surveillance is expected around 2027; NHS England Digital updates its autism waiting-time statistics quarterly; the Global Burden of Disease Study refreshes on a multi-year cycle (a GBD 2023 update is anticipated); and the WHO periodically revisits its global estimate. The Dietz adult estimate and the Buescher lifetime-cost figures are the most recent authoritative US numbers of their kind and are candidates for update.
Last updated: July 17, 2026.
We review and update this page quarterly as new data is published.