By age 4, 11.2% of children in a community cohort of 1,619 Melbourne children had started to stutter, yet only 6.3% of them recovered within the first 12 months (Reilly et al., Pediatrics 2013). Roughly 3 million Americans stutter, according to the NIH’s National Institute on Deafness and Other Communication Disorders, and the Stuttering Foundation puts the worldwide count at more than 80 million people. The condition is also far more genetic than once assumed: a 2025 Nature Genetics study of more than one million people identified 57 unique loci linked to stuttering. This roundup focuses on who starts stuttering, who keeps stuttering, the mental health picture and what the treatment trials show; our broader stuttering statistics post covers workplace outcomes. We aggregated data from NIDCD, the CDC’s National Center for Health Statistics, the Stuttering Foundation, the National Stuttering Association, the Cochrane Library, the BMJ, Pediatrics, Nature Genetics, the Journal of Fluency Disorders and the other peer-reviewed sources listed in the methodology.
TL;DR
- Roughly 3 million Americans stutter; 5 to 10% of all children stutter for some period (NIDCD, Stuttering 2017).
- More than 80 million people worldwide stutter, about 1% of the population (Stuttering Foundation, FAQ).
- Cumulative incidence of stuttering onset by age 4: 11.2% (Reilly et al., Pediatrics 2013).
- Approximately 75% of children who stutter recover (NIDCD, Stuttering 2017).
- Male-to-female ratio rises from 2.3:1 in young children to 4:1 in adolescence (Craig et al., JSLHR 2002).
- A GWAS of 99,776 cases and 1,023,243 controls found 57 unique loci (Polikowsky et al., Nature Genetics 2025).
- Adults seeking stuttering treatment had 16- to 34-fold increased odds of social phobia (Iverach et al., Journal of Anxiety Disorders 2009).
- 38.4% of children who stutter had at least one mental health concern vs 13.1% of peers (Merlo and Briley, Journal of Fluency Disorders 2025).
- Lidcombe Program: 1.5% vs 3.9% syllables stuttered at nine months, treatment vs control (Jones et al., BMJ 2005).
- 76.5% of Lidcombe and 71.4% of RESTART-DCM children were not stuttering at 18 months (de Sonneville-Koedoot et al., PLOS ONE 2015).
- Telehealth vs in-person therapy: no significant difference in stuttering frequency up to 18 months across 9 RCTs (Scott et al., Journal of Telemedicine and Telecare 2025).
- 67.6% of US children with speech problems received intervention services (NCHS Data Brief 205, 2012 data).
1. How Many People Stutter: Prevalence Across the Lifespan
Every headline figure for stuttering rests on two different questions: how many people stutter right now (prevalence) and how many ever have (incidence). The gap between them is the story of the condition. Roughly 3 million Americans stutter, and approximately 5 to 10% of all children will stutter for some period in their life (NIDCD, Stuttering). The Stuttering Foundation puts the global figure at more than 80 million people, about 1% of the population, and the National Stuttering Association uses 79 million worldwide.
Careful population studies come in lower than the popular 1% rule. An Australian household survey that confirmed cases from recorded speech found a whole-population prevalence of 0.72%, highest in young children (1.4 to 1.44%) and lowest in adolescence (0.53%), with lifetime incidence of 2.1% in adults aged 21 to 50 (Craig et al., Journal of Speech, Language, and Hearing Research 2002). A major review concluded that lifespan incidence may be higher than the 5% commonly cited while average lifespan prevalence may be lower than 1% (Yairi and Ambrose, Journal of Fluency Disorders 2013). In other words, many more people stutter at some point than the 1% figure implies, but far fewer keep stuttering.
| Metric | Value | Source |
|---|---|---|
| Americans who stutter | Roughly 3 million | NIDCD, Stuttering (updated 2017) |
| Children who stutter for some period | Approximately 5 to 10% | NIDCD, Stuttering (updated 2017) |
| People who stutter worldwide | More than 80 million, about 1% of the population | Stuttering Foundation, FAQ |
| People who stutter worldwide (NSA figure) | 79 million | National Stuttering Association, homepage |
| Whole-population prevalence, New South Wales | 0.72% (young children 1.4 to 1.44%, adolescents 0.53%) | Craig et al., JSLHR 2002 |
| Lifetime incidence by age group | 2.1% adults 21-50; 2.8% ages 2-5; 3.4% ages 6-10 | Craig et al., JSLHR 2002 |
| Preschool vs adult burden | More than 5% of preschoolers; chronic in about 1% of adults | Chang et al., Neurobiology of Language 2025 |
| School-age prevalence, Tamil Nadu, India | 0.46% of 74,544 children screened | Nandhini Devi et al., Journal of Fluency Disorders 2018 |
Context note: the spread between 0.46% in Indian schoolchildren and 0.72% in an Australian household survey reflects case definitions and ascertainment methods as much as real differences. Yairi and Ambrose state that the effects of race, ethnicity, culture, bilingualism and socioeconomic status on stuttering incidence and prevalence remain uncertain.
2. Onset and Natural Recovery: A Preschool Condition That Mostly Resolves
Stuttering is overwhelmingly a condition that begins in early childhood, and the newest cohort data push the onset window even earlier. In the Early Language in Victoria Study, cumulative incidence of stuttering onset reached 11.2% (95% CI 9.7% to 12.8%) by age 4 (Reilly et al., Pediatrics 2013). That is more than double the 5% figure long cited in textbooks. Higher maternal education, male gender and twinning predicted onset, and at age 4 the children who stuttered actually had stronger language scores (105.0 vs 99.6) than their non-stuttering peers.
Recovery is the norm, but it is slow. Only 9 of 142 children (6.3%) in the Melbourne cohort recovered within 12 months of onset, which is why NIDCD notes that health professionals recommend evaluation for a child who has stuttered for 3 to 6 months. A 2025 study of 2,055 three-year-olds in Japan found 6.5% stuttering at their health checkup and 8.9% once past stuttering was included (Sakai et al., Folia Phoniatrica et Logopaedica 2025). For context on how many of these children end up in a clinic, see our speech therapy statistics.
| Metric | Value | Source |
|---|---|---|
| Cumulative incidence of onset by age 4 | 11.2% (95% CI 9.7% to 12.8%) | Reilly et al., Pediatrics 2013 |
| Recovered within 12 months of onset | 9 of 142 (6.3%) | Reilly et al., Pediatrics 2013 |
| Peak onset ages | Between ages 2 and 6 | NIDCD, Stuttering (updated 2017) |
| Children who recover | Approximately 75%; 25% may persist lifelong | NIDCD, Stuttering (updated 2017) |
| Children stuttering 6+ months who recover by late childhood | Three-quarters, leaving about 1% with a long-term problem | Stuttering Foundation, FAQ |
| Stuttering at age-3 checkup, Japan | 6.5% (8.9% including past stuttering), n = 2,055 | Sakai et al., Folia Phoniatrica et Logopaedica 2025 |
| Persistent stuttering class, low-income US children | 3.5% (typical speech 85%, other speech problems 11.5%) | Liu et al., JSLHR 2026 |
| Comorbid speech sound disorder at age 4 | 6.88% of children who stutter | Unicomb et al., Journal of Communication Disorders 2020 |
Context note: Yairi and Ambrose concluded that most of the risk of onset is over by age 5 and that the male-to-female ratio near onset is smaller than previously thought. The NIDCD page was last updated in March 2017, so its figures are the most recent available from that source.
3. Sex, Family History and Genetics: Who Keeps Stuttering
The demographic split widens with age because girls recover more often than boys. NIDCD reports that boys are 2 to 3 times as likely to stutter as girls, and the number of boys who continue to stutter is three to four times larger than the number of girls (NIDCD, Stuttering). The Australian community data show the same drift, from a 2.3:1 male-to-female ratio in younger children to 4:1 in adolescence.
Family history is the strongest single risk signal, and genetics has moved from family trees to genome-wide scans. A 2025 meta-analysis of more than one million people (99,776 cases and 1,023,243 controls) identified 57 unique loci and found genetic similarity between stuttering and autism, depression and impaired musical rhythm (Polikowsky et al., Nature Genetics 2025). The overlap with autism is worth reading alongside our autism prevalence statistics.
| Metric | Value | Source |
|---|---|---|
| Boys vs girls who stutter | 2 to 3 times as likely; 3 to 4 times as many boys persist | NIDCD, Stuttering (updated 2017) |
| Male-to-female ratio by age | 2.3:1 young children, 4:1 adolescence, 2.3:1 all ages | Craig et al., JSLHR 2002 |
| Male-to-female ratio, Tamil Nadu schoolchildren | 4:1; positive family history in 56% of probands | Nandhini Devi et al., Journal of Fluency Disorders 2018 |
| People who stutter with a family member who stutters | Approximately 60% | Stuttering Foundation, FAQ |
| Odds of stuttering with family history, Japan | OR 3.27 | Sakai et al., Folia Phoniatrica et Logopaedica 2025 |
| Reported positive family history across 19 studies | 20.0% to 90.9% | Sato et al., International Journal of Language and Communication Disorders 2026 |
| Largest published GWAS | 99,776 cases, 1,023,243 controls, 57 unique loci | Polikowsky et al., Nature Genetics 2025 |
| Stuttering in people with Down syndrome | 2.38% to 56% across 14 studies | Hokstad and Naess, Frontiers in Psychology 2023 |
Context note: two 2026 preprints extend the genetic picture but are not yet peer reviewed. One reports 150 unique causal stuttering genes from a GWAS of 78,394 cases and 865,956 controls; another, a meta-analysis of 18 clinical cohorts (6,096 cases, 81,629 controls), estimates SNP-based heritability at about 0.26 and links suggestive loci to childhood apraxia of speech genes (medRxiv, Scartozzi et al. and Jackson et al., 2026).
4. Mental Health and Co-occurring Conditions
The fluency problem is only part of the burden. Adults seeking speech treatment for stuttering had six- to seven-fold increased odds of a 12-month anxiety disorder diagnosis and 16- to 34-fold increased odds of social phobia (Iverach et al., Journal of Anxiety Disorders 2009). The same research group found twofold odds of a mood disorder in treatment-seeking adults (2010) and, in children, six-fold odds of social anxiety disorder and four-fold odds of any anxiety disorder (2016).
Population survey data confirm that this is not only a clinic effect. Among 155 children who stutter in the 2012 National Health Interview Survey, 38.4% had at least one mental health concern, versus 13.1% of 8,547 controls, and 28.0% had at least one sleep problem versus 11.4% (Merlo and Briley, Journal of Fluency Disorders 2025). The same national survey shows adults who stutter with roughly double the odds of asthma and allergies. Anxiety and avoidance can also affect the voice itself, a link explored in our voice disorders statistics.
| Metric | Value | Source |
|---|---|---|
| Odds of any anxiety disorder, treatment-seeking adults | Six- to seven-fold; social phobia 16- to 34-fold | Iverach et al., Journal of Anxiety Disorders 2009 |
| Odds of any mood disorder, treatment-seeking adults | Twofold | Iverach et al., JSLHR 2010 |
| Children who stutter: anxiety odds vs controls | Six-fold social anxiety disorder; four-fold any anxiety disorder | Iverach et al., Journal of Fluency Disorders 2016 |
| Adolescents with at least one mental disorder diagnosis | 38%, double general adolescent estimates | Gunn et al., Journal of Fluency Disorders 2014 |
| Children with at least one mental health concern (NHIS 2012) | 38.4% vs 13.1% of controls | Merlo and Briley, Journal of Fluency Disorders 2025 |
| Children with at least one sleep problem (NHIS 2012) | 28.0% vs 11.4% of controls | Merlo and Briley, Journal of Fluency Disorders 2025 |
| Children who stutter diagnosed with ADHD | 17.2% of 204 children | Walsh et al., JSLHR 2025 |
| Adults who stutter vs controls (NHIS 2012) | Asthma OR 2.30; respiratory allergies OR 2.38; excessive sleepiness OR 2.67 | Merlo and Briley, Journal of Fluency Disorders 2024 |
Context note: anxiety estimates from treatment-seeking samples overstate what a random person who stutters experiences. In a separate 2011 scale-development study, approximately one-quarter of participants who stutter met criteria for social phobia (23.5% under DSM-IV and 27.2% under ICD-10; Iverach et al., 2011). The Melbourne cohort found no worse health-related quality of life in the first year after onset, so mental health effects appear to build over time rather than start at onset.
5. Therapy Outcomes for Preschool Children: What the Trials Show
Early intervention has the strongest evidence base in the field, almost all of it for one program. In the first randomised trial of the Lidcombe Program, treated preschoolers stuttered on 1.5% of syllables at nine months versus 3.9% in the control arm, an effect size of 2.3 percentage points (Jones et al., BMJ 2005). The Dutch RESTART trial then compared Lidcombe with an indirect approach (RESTART-DCM) in 199 children and found that direct treatment worked faster in the first three months but outcomes converged by 18 months (de Sonneville-Koedoot et al., PLOS ONE 2015).
The systematic reviews are more cautious than the individual trials. The Cochrane review pooled four Lidcombe-vs-wait-list RCTs with 151 children aged 2 to 6 and found a mean difference of -2.16 percentage points of syllables stuttered, but rated the evidence very low certainty, partly because no child had finished the full program, which typically takes 1 to 2 years (Sjostrand et al., Cochrane Database of Systematic Reviews 2021). A review of 8 RCTs found no high-level evidence for school-aged children or adolescents.
| Metric | Value | Source |
|---|---|---|
| Syllables stuttered at 9 months, Lidcombe vs control | 1.5% vs 3.9% (effect size 2.3%, 95% CI 0.8 to 3.9), n = 54 | Jones et al., BMJ 2005 |
| Not stuttering at 18 months | Lidcombe 76.5% (65/85) vs RESTART-DCM 71.4% (65/91) | de Sonneville-Koedoot et al., PLOS ONE 2015 |
| Syllables stuttered at 18 months | Lidcombe 1.2% vs RESTART-DCM 1.5% | de Sonneville-Koedoot et al., PLOS ONE 2015 |
| Cochrane pooled effect, 4 RCTs, 151 children | MD -2.16 (95% CI -3.48 to -0.84), very low certainty | Cochrane Review CD013489, 2021 |
| Pooled Lidcombe effect, 8-RCT review | -3.8 (95% CI -7.3 to -0.3) | Brignell et al., Journal of Fluency Disorders 2021 |
| USPSTF evidence report, 2 Lidcombe RCTs (n = 76) | 2.3% to 3.0% lower syllables stuttered at 9 months | Feltner et al., JAMA 2024 |
| Community clinic translation | 51 clinicians, 121 children; outcomes similar to RCTs | O’Brian et al., International Journal of Language and Communication Disorders 2023 |
Context note: the RESTART economic evaluation found differences in effects and costs between the two programs were small, with cost-effectiveness ratios in favor of Lidcombe (Journal of Communication Disorders, 2015). The Lidcombe evidence base is dominated by its developers’ research group, and the 249-child TreatPaCS non-inferiority trial (protocol, PLOS ONE 2024) will add an independent three-way comparison.
6. Older Children, Adults, Telehealth and Access to Care
Outside the preschool window the evidence thins out quickly. A Phase II telehealth trial of the Lidcombe Program for 37 children aged 6 to 12 found only 7 (18.9%) met Stage 2 criteria at six months, while 25 (67.6%) showed a partial response and 5 (13.5%) no response (Johnson et al., Journal of Fluency Disorders 2024). Drug treatment is no shortcut: a 2025 review of 39 studies covering 17 drug classes rated only 7 of 19 RCTs at low or some risk of bias (Horton et al., Neuroscience and Biobehavioral Reviews 2025), and a meta-analysis found delayed auditory feedback alone gave no significant overall benefit.
Delivery format matters less than access. A meta-analysis of 9 RCTs found no significant difference in stuttering frequency between telehealth and face-to-face therapy at any time point up to 18 months (Scott et al., Journal of Telemedicine and Telecare 2025). The access gap is the bigger problem: in national CDC data, 67.6% of US children with speech problems, a category that explicitly includes stuttering, received intervention services, and only 45.8% of non-Hispanic Black children with any communication disorder did, versus 60.1% of non-Hispanic White children (NCHS Data Brief No. 205).
| Metric | Value | Source |
|---|---|---|
| School-age Lidcombe telehealth, 6-month response (n = 37) | 18.9% full, 67.6% partial, 13.5% none | Johnson et al., Journal of Fluency Disorders 2024 |
| Telehealth vs in-person, % syllables stuttered (9 RCTs) | MD 0.1 (95% CI -0.4 to 0.6), p = 0.70 | Scott et al., Journal of Telemedicine and Telecare 2025 |
| Webcam Lidcombe noninferiority RCT | 49 children; insufficient evidence of a posttreatment difference | Bridgman et al., JSLHR 2016 |
| Pharmacological studies reviewed | 39 studies, 17 drug classes; 87% in adolescents/adults; 7 of 19 RCTs low/some risk of bias | Horton et al., Neuroscience and Biobehavioral Reviews 2025 |
| Delayed auditory feedback vs normal feedback | MD -1.46 (95% CI -4.83 to 1.91), not significant; 5 studies, 61 participants | Iimura et al., IJLCD 2026 |
| Adult relapse over 24 months after treatment | 1 of 19 (5.3%) sustained relapse | Almudhi, Healthcare 2026 |
| US children 3-17 with speech problems | 5.0%; 67.6% of them received services | NCHS Data Brief 205 (2012 NHIS) |
| Services received, any communication disorder, by race | Non-Hispanic White 60.1%, Hispanic 47.3%, non-Hispanic Black 45.8% | NCHS Data Brief 205 (2012 NHIS) |
Context note: clinician confidence is part of the access problem. A systematic review of 13 studies found speech-language pathologists rate their competence lower for treating stuttering than for assessing it, and higher when they see people who stutter more often (Alegre et al., IJLCD 2025). The NCHS brief uses 2012 data and remains the most recent federal breakdown of this kind.
Summary: Stuttering and Speech Dysfluency by the Numbers
| Metric | Value | Source |
|---|---|---|
| Americans who stutter | Roughly 3 million | NIDCD, Stuttering (2017) |
| People who stutter worldwide | More than 80 million (about 1%) | Stuttering Foundation, FAQ |
| Children who stutter for some period | 5 to 10% | NIDCD, Stuttering (2017) |
| Whole-population prevalence, Australia | 0.72% | Craig et al., JSLHR 2002 |
| Cumulative incidence by age 4 | 11.2% | Reilly et al., Pediatrics 2013 |
| Recovered within 12 months of onset | 6.3% | Reilly et al., Pediatrics 2013 |
| Children who eventually recover | Approximately 75% | NIDCD, Stuttering (2017) |
| Stuttering at age-3 checkup, Japan | 6.5% (8.9% incl. past) | Sakai et al., 2025 |
| Male-to-female ratio, adolescence | 4:1 | Craig et al., JSLHR 2002 |
| Family member who stutters | About 60% | Stuttering Foundation, FAQ |
| Genetic loci identified | 57 (99,776 cases) | Polikowsky et al., Nature Genetics 2025 |
| Social phobia odds, treatment-seeking adults | 16- to 34-fold | Iverach et al., 2009 |
| Adolescents with a mental disorder diagnosis | 38% | Gunn et al., 2014 |
| Children with a mental health concern | 38.4% vs 13.1% | Merlo and Briley, 2025 |
| Lidcombe vs control at 9 months | 1.5% vs 3.9% syllables stuttered | Jones et al., BMJ 2005 |
| Not stuttering at 18 months | 76.5% Lidcombe, 71.4% RESTART-DCM | de Sonneville-Koedoot et al., 2015 |
| Cochrane pooled effect | MD -2.16, very low certainty | Cochrane Review CD013489, 2021 |
| Telehealth vs in-person | No significant difference (9 RCTs) | Scott et al., 2025 |
| Children with speech problems receiving services | 67.6% | NCHS Data Brief 205 |
| School-age Lidcombe full responders | 18.9% | Johnson et al., 2024 |
Methodology and Sources
Every figure above was read during research for this article on a government health page, an organization’s published page, a trial publication or a peer-reviewed abstract retrieved through the Europe PMC database. Statistics that circulate on blogs without a primary study were excluded, and advocacy-organization figures (Stuttering Foundation, National Stuttering Association) are labeled as such. This article is informational and is not medical advice.
- NIH, National Institute on Deafness and Other Communication Disorders: Stuttering (last updated March 2017)
- CDC, National Center for Health Statistics: Data Brief No. 205, Communication Disorders and Use of Intervention Services Among Children Aged 3-17 Years (June 2015, 2012 NHIS data)
- Stuttering Foundation: Frequently Asked Questions
- National Stuttering Association: westutter.org
- Yairi and Ambrose: Epidemiology of stuttering: 21st century advances, Journal of Fluency Disorders 2013
- Craig et al.: Epidemiology of stuttering in the community across the entire life span, JSLHR 2002
- Reilly et al.: Natural history of stuttering to 4 years of age, Pediatrics 2013
- Sakai et al.: Multiple-community-based epidemiological study of stuttering among 3-year-old children in Japan, Folia Phoniatrica et Logopaedica 2025
- Polikowsky et al.: Large-scale genome-wide analyses of stuttering, Nature Genetics 2025
- Iverach et al.: Prevalence of anxiety disorders among adults seeking speech therapy for stuttering, Journal of Anxiety Disorders 2009
- Merlo and Briley: Influence of mental health on sleep in children who stutter, Journal of Fluency Disorders 2025
- Jones et al.: Randomised controlled trial of the Lidcombe programme of early stuttering intervention, BMJ 2005
- de Sonneville-Koedoot et al.: Direct versus indirect treatment for preschool children who stutter: the RESTART randomized trial, PLOS ONE 2015
- Sjostrand et al.: Non-pharmacological interventions for stuttering in children six years and younger, Cochrane Database of Systematic Reviews 2021
- Johnson et al.: Lidcombe Program telehealth treatment for children 6-12 years of age, a Phase II trial, Journal of Fluency Disorders 2024
- Horton et al.: Pharmacological and dietary treatments for developmental stuttering, a systematic review, Neuroscience and Biobehavioral Reviews 2025
- Scott et al.: Telehealth versus face-to-face delivery of speech language pathology services, a systematic review and meta-analysis, Journal of Telemedicine and Telecare 2025
- Additional peer-reviewed abstracts read via Europe PMC and cited by author and year in the tables: Chang et al. (Neurobiology of Language 2025), Nandhini Devi et al. (Journal of Fluency Disorders 2018), Liu et al. (JSLHR 2026), Unicomb et al. (Journal of Communication Disorders 2020), Sato et al. (IJLCD 2026), Hokstad and Naess (Frontiers in Psychology 2023), Iverach et al. (JSLHR 2010; Journal of Fluency Disorders 2016; IJLCD 2011), Gunn et al. (Journal of Fluency Disorders 2014), Walsh et al. (JSLHR 2025), Merlo and Briley (Journal of Fluency Disorders 2024), Brignell et al. (Journal of Fluency Disorders 2021), Feltner et al. (JAMA 2024), O’Brian et al. (IJLCD 2023), Bridgman et al. (JSLHR 2016), Iimura et al. (IJLCD 2026), Almudhi (Healthcare 2026), Alegre et al. (IJLCD 2025), and the medRxiv preprints by Scartozzi et al. and Jackson et al. (2026).
- Data watch: much of the core epidemiology is old. The NIDCD page was last updated in 2017, the NCHS brief uses 2012 NHIS data, the Craig survey dates from 2002 and the first Lidcombe RCT from 2005; these are labeled as the most recent available data from each source. Prevalence estimates conflict: the Stuttering Foundation’s ‘about 1%’ and the NSA’s 79 million worldwide are advocacy figures, while population studies found 0.72% (Australia) and 0.46% (Tamil Nadu schoolchildren), and Yairi and Ambrose suggest lifespan prevalence may be below 1%. Incidence also conflicts: NIDCD cites 5 to 10% of children, the Stuttering Foundation about 5% for six months or more, and the Melbourne cohort 11.2% by age 4. The NCHS ‘speech problems’ category includes stuttering but also speech sound problems, so 5.0% is not a stuttering prevalence. Anxiety odds come from treatment-seeking samples and should not be read as population rates. The 150-gene and h2 0.26 figures are preprints not yet peer reviewed. The ASHA Practice Portal page on fluency disorders returned 403 and could not be read, so no ASHA figures are used. Several PubMed pages were unreadable; abstracts were read through the Europe PMC API and linked by DOI.
Last updated: October 3, 2026. We update this roundup quarterly, and the next refresh is expected when the TreatPaCS trial reports its 18-month results and the Nature Genetics follow-up preprints from the Below lab complete peer review.