About two-thirds of young American women in a landmark Journal of Voice sample used vocal fry, yet when 800 listeners heard young women’s voices in fry, they judged them less competent, less educated, less trustworthy, less attractive and less hirable (Wolk, Abdelli-Beruh and Slavin, Journal of Voice 2012; Anderson et al., PLOS ONE 2014). Those two findings defined a decade of headlines, but the research base behind them is thin: a 2020 systematic review screened 2,021 records and found only ten studies that measured creaky voice prevalence in English, covering 297 speakers in total. Newer work points in different directions: Finnish and Australian archive studies now show a real increase in creak among young women, while a 2026 study found no overall hirability penalty for natural phrase-final creak. Voice quality is only one piece of how listeners size up a speaker. We aggregated data from the Journal of Voice, PLOS ONE, JASA Express Letters, Language and Speech, Folia Phoniatrica et Logopaedica, Logopedics Phoniatrics Vocology and the other peer-reviewed sources listed in the methodology.
TL;DR
- Approximately two-thirds of 34 female US college students used vocal fry, mostly sentence-finally (Wolk et al., Journal of Voice 2012).
- 86% of 50 female undergraduates produced at least one episode of fry even in isolated nonwords (Gibson, Journal of Voice 2017).
- Only ten studies worldwide met inclusion criteria for measuring creak prevalence in English, with 297 speakers in total (Dallaston and Docherty, PLOS ONE 2020).
- Finnish women rated as using ‘quite a lot’ of creak rose from 7% to 25.8% between the 1990s and 2010s (Uusitalo et al., Journal of Voice 2024).
- Sydney teenage girls showed an increase in creak over 30 years; boys did not (White, Penney and Cox, JASA Express Letters 2025).
- 800 listeners preferred normal voices over fry at rates significantly better than chance (p < .001) on all five judgments (Anderson et al., PLOS ONE 2014).
- Female speakers lost on attractiveness and intelligence with fry; male speakers did not (Taylor et al., Journal of Voice 2025).
- 104 listeners did not rate spontaneous phrase-final creak as less hirable overall (Bellavance and Levi, Language and Speech 2026).
- NPR’s female broadcasters used brief fry in 22% of 267 archival samples across six decades (Stewart et al., Journal of Voice 2025).
- Bilingual native English speakers produced fry 4.6% of the time in English vs 2.9% in Spanish (Cantor-Cutiva et al., Journal of Voice 2023).
- Creak appeared in 22% of breath groups for laryngeal dystonia patients vs 5% for controls (Marks et al., Journal of Voice 2024).
1. How Common Is Vocal Fry? Prevalence in American English
The popular claim that ‘young women all talk in fry’ traces back to very small samples, but those samples are consistent. Approximately two-thirds of 34 female college students aged 18-25 used vocal fry, and it was most likely to occur at the end of sentences (Wolk, Abdelli-Beruh and Slavin, Journal of Voice 2012). When the same lab recorded 34 matched male students, fry prevalence in sentences was significantly lower than in the women, and in both sexes fry clustered in sentence-final position. A later study stripped away syntax and meaning entirely: 86% of 50 female undergraduates still produced at least one episode of fry while repeating nonwords, more often on unstressed than stressed syllables (Gibson, Journal of Voice 2017).
What these numbers mean is that fry behaves less like a fashion statement and more like a default prosodic habit: it shows up where pitch falls at the end of a phrase and disappears when stress or volume pushes pitch up. Self-perception lags behind behavior. In a 40-student study, only 18% of women and 6% of men said they used vocal fry, even though three speech-language pathologists identified a high occurrence of fry across nine simulated room conditions (Cantor-Cutiva et al., Logopedics Phoniatrics Vocology 2018).
| Metric | Value | Source |
|---|---|---|
| Female US college students using fry (n = 34, ages 18-25) | Approximately two-thirds | Wolk et al., Journal of Voice 2012 |
| Female undergraduates producing at least one fry episode in nonwords (n = 50) | 86% | Gibson, Journal of Voice 2017 |
| Male college students, fry in sentences (n = 34) | Significantly lower than matched female sample | Abdelli-Beruh et al., Journal of Voice 2014 |
| Creaky words, American women vs men (Yuasa 2010, 12 women, 11 men) | About 20% vs about 10% of words | Dallaston and Docherty, PLOS ONE 2020 |
| Self-reported fry use, college students (22 women, 18 men) | 18% of women (4 of 22), 6% of men (1 of 18) | Cantor-Cutiva et al., Logopedics Phoniatrics Vocology 2018 |
| Odds of fry with sporadic caffeine use (5 or fewer per week) | OR 0.3 (95% CI 0.1-0.9) | Cantor-Cutiva et al., Logopedics Phoniatrics Vocology 2018 |
| Odds of fry in babble noise / pink noise | OR 0.5 (0.4-0.8) / OR 0.4 (0.2-0.8) | Cantor-Cutiva et al., Logopedics Phoniatrics Vocology 2018 |
| Automated fry detection, 25 men and 29 women reading | Fry frequencies greatest at sentence end, unaffected by sentence length | Abdelli-Beruh et al., Journal of Voice 2016 |
Context note: the Yuasa (2010, American Speech) word-level figures are approximate values read from graphs by the systematic review authors, not exact published percentages. Most recent available data for the core US prevalence estimates: Wolk et al. 2012 and Gibson 2017; no large, representative US prevalence survey exists.
2. Is Creaky Voice Rising? The New Over-Time Evidence
For years the honest answer was ‘nobody knows’. A 2020 systematic review screened 2,021 titles and abstracts, read 785 full texts and found only ten studies that met its criteria, concluding that widespread claims of a recent increase among young American women have not been empirically confirmed (Dallaston and Docherty, PLOS ONE 2020). Nine of the ten studies analyzed American English, 64% of the 297 speakers were female, and at least 82 of the 149 American women were college students under 30, so the literature mostly measured one demographic at one moment.
That gap is now closing with archive recordings. A Finnish study compared 200 university students recorded in the 1990s and 2010s and found mean perceived creak in women rose from 1.04 to 1.55 on a 0-4 scale, with the share rated ‘quite a lot’ climbing from 7% to 25.8% (Uusitalo et al., Journal of Voice 2024). In Sydney, a real-time comparison of 28 teenagers recorded in 1989 with 21 teenagers recorded 30 years later found an increase in creak prevalence for girls but not boys (White, Penney and Cox, JASA Express Letters 2025). Professional voices moved less: NPR’s female broadcasters showed brief fry in 22% of samples, with glottal fry per minute ratios of only 0.47-0.86, markedly lower than reported norms for young and middle-aged women (Stewart et al., Journal of Voice 2025).
| Metric | Value | Source |
|---|---|---|
| Studies meeting inclusion criteria for English creak prevalence | 10 (from 3,990 records, 785 full-text reviews) | Dallaston and Docherty, PLOS ONE 2020 |
| Speakers across those 10 studies | 297 (64% female, n = 189) | Dallaston and Docherty, PLOS ONE 2020 |
| Finnish women rated ‘quite a lot’ of creak, 1990s vs 2010s | 7% to 25.8% (‘a lot’: 0% to 6%) | Uusitalo et al., Journal of Voice 2024 |
| Finnish men rated ‘quite a lot’ of creak, 1990s vs 2010s | 5.9% to 18.1% (‘a lot’: 0% to 2.3%) | Uusitalo et al., Journal of Voice 2024 |
| Mean perceived creak in Finnish women (0-4 scale) | 1.04 to 1.55 (p < 0.05); no significant change in men | Uusitalo et al., Journal of Voice 2024 |
| Sydney teens, 1989 (n = 28) vs 30 years later (n = 21) | Creak increased for girls, not boys | White, Penney and Cox, JASA Express Letters 2025 |
| NPR female broadcasters, samples with brief fry | 22% of 267 samples (18 broadcasters, 1970s-2020s) | Stewart et al., Journal of Voice 2025 |
| NPR broadcasters’ mean speaking pitch | 188 Hz overall; 196 Hz (2000s) high, 176 Hz (2010s) low | Stewart et al., Journal of Voice 2025 |
Context note: the direction of the gender gap is not fixed. In British recordings from the mid-1980s, Modified Northern men produced creaky syllables at a rate approximately 6 times higher than women (Henton and Bladon 1988, as summarized in Dallaston and Docherty 2020). In the Finnish data, age did not correlate with creak, which points to a generational shift rather than an age effect.
3. The Listener Penalty: Hiring, Competence and Attractiveness
The most quoted vocal fry number is not a prevalence figure but a perception one. In a sample of 800 listeners (400 men, 400 women, ages 18-65), preference ratios for fry fell substantially below .50 on all five judgments, meaning listeners chose the normal voice as more trustworthy, competent, educated, attractive and hirable at rates significantly better than chance (p < .001) (Anderson et al., PLOS ONE 2014). The 14 speakers each recorded the same job-interview line, ‘Thank you for considering me for this opportunity’, in both voices.
The headline penalty is real but modest in effect-size terms, and it lands harder on women. In Anderson’s data the sex-of-speaker effect for women on competence had a partial eta squared of 1.26%. A 2025 replication with 29 listeners found female speakers without fry were rated significantly more attractive and intelligent than the same speakers with fry, while ratings of male speakers did not move (Taylor et al., Journal of Voice 2025). The same asymmetry shows up in research on how accents affect hiring decisions: small acoustic cues trigger large social inferences.
| Metric | Value | Source |
|---|---|---|
| Listeners in hiring-voice study | 800 (400 men, 400 women; ages 18-33, 34-50, 51-65) | Anderson et al., PLOS ONE 2014 |
| Preference for normal voice over fry | Significantly better than chance on all 5 judgments (p < .001) | Anderson et al., PLOS ONE 2014 |
| Female-speaker effect size, competence / trustworthiness | Partial eta squared 1.26% / 1.19% | Anderson et al., PLOS ONE 2014 |
| Overall effect of fry on attribute ratings (29 listeners) | Partial eta squared 0.393 (p = 0.005) | Taylor et al., Journal of Voice 2025 |
| Fry x gender interaction, attractiveness / intelligence | Partial eta squared 0.304 / 0.251; male speakers unaffected | Taylor et al., Journal of Voice 2025 |
| Listeners judging intelligence and likability (8 speakers) | 463 adults; fry favorable or unfavorable depending on pitch and rate | Parker et al., Journal of Voice 2018 |
| Simulated continuous fry, 30 minutes (10 speakers, 10 listeners) | Rated less employable, less natural, needing more concentration | Venkatraman et al., American Journal of Speech-Language Pathology 2018 |
Context note: Anderson’s speakers produced fry on purpose, and the 2026 Bellavance and Levi study was designed specifically to test spontaneously produced creak instead. Parker et al.’s 463-listener study found the same fry could help in one context (high pitch, fast rate) and hurt in another (low pitch, fast rate), which is a strong argument against treating fry as a single on-off switch. Speaking tempo matters too; see our speech rate statistics.
4. Who Is Listening: Age, Gender and Context Reverse the Verdict
The newest perception studies suggest the penalty depends on the listener at least as much as the speaker. When 104 listeners heard spontaneously produced phrase-final creak from 10 speakers, they did not rate it as less hirable or less pleasant overall; only older women rated creaky sentences as less pleasant (Bellavance and Levi, Language and Speech 2026). In New Zealand, 60 listeners split by age showed the same fault line: younger listeners rated creak as equally pleasant, educated and confident as a neutral voice, while older listeners rated it significantly more annoying, ignorant and unsure (Calhoun and White, Language and Speech 2025).
The implication is generational. Creak looks like a feature whose social meaning is being renegotiated in real time, with the evaluators who make hiring decisions today skewing older than the speakers who use it most. Even trained ears are not neutral: in one study, graduate students in communication disorders linked fry most often to ‘vain’ and ‘apathetic’, and in another, a similar group spotted fry with 85% accuracy but labeled ordinary modal voice correctly only 51% of the time.
| Metric | Value | Source |
|---|---|---|
| Listeners rating spontaneous phrase-final creak | 104 listeners, 5 groups, 10 speakers; no overall hirability or pleasantness penalty | Bellavance and Levi, Language and Speech 2026 |
| Mean affective rating, neutral vs creak vs uptalk (60 NZ listeners) | 54.9 vs 51.8 vs 40.8 | Calhoun and White, Language and Speech 2025 |
| Creak rating, older (50-75) vs younger (18-30) listeners | 50.2 vs 53.5; older harsher on pleasantness (p = .0045), education (p = .0189), certainty (p = .0228) | Calhoun and White, Language and Speech 2025 |
| Accuracy identifying fry / uptalk / modal voice (graduate students) | 85% / 79% / 51% | Stewart, Kling and D’Agosto, Journal of Voice 2026 |
| Graduate students describing fry only with undesirable adjectives (n = 23) | 32% (64% mixed, 5% only desirable) | Ligon et al., Journal of Voice 2019 |
| Adjectives most linked to fry | Vain 56%, apathetic 48%, sleepy 38%, relaxed 38%, bored 36% | Ligon et al., Journal of Voice 2019 |
| Desirability score rank of fry among 8 voice qualities | 4th, at 1.82 (low pitch 2.13 highest, strained 1.08 lowest) | Ligon et al., Journal of Voice 2019 |
| Listeners rating fry in a cappella singing | 253 participants; female singers with fry rated more sincere/committed, male singers less | Chappell, Journal of Voice 2020 |
Context note: in the a cappella study, younger and less musically trained listeners evaluated fry more positively, and listeners with more musical training rated singers with fry as less confident (Chappell, Journal of Voice 2020). An identification study in Phonetica found listeners were least accurate at spotting partial, utterance-final creak and showed no strong gender bias in detecting it (Davidson, Phonetica 2019).
5. Beyond American English, and Where Creak Becomes Clinical
Creak is not an American invention, and its rate depends on language exposure, speaking style and, in some speakers, laryngeal health. Bilingual speakers who were native English speakers produced vocal fry 4.6% of the time in English versus 2.9% in Spanish, while native Spanish speakers produced 1.9% versus 0.9% (Cantor-Cutiva et al., Journal of Voice 2023). An El Paso study of 58 women across monolingual English, monolingual Spanish and bilingual groups found no significant group differences in fry frequency, but all groups produced more fry when repeating English nonwords, which its authors read as evidence that mere exposure to English, not sociolinguistic signaling, drives much of the habit (Gibson et al., Journal of Voice 2017). In Dutch men, creaky syllables ranged roughly from 0% to 5% per speaker across at least 1,500 syllables each (van Hugte et al., Journal of Voice 2026).
At the clinical end, creak is becoming a measurable biomarker rather than a style. In adductor laryngeal dystonia, creak appeared in 22% of breath groups versus 5% in typical speakers (Marks et al., Journal of Voice 2024), and automated creak scores now separate patients from controls with areas under the curve up to 0.88. That distinction matters because ordinary sentence-final fry is not, on current evidence, a disorder; the background on genuine voice pathology is in our voice disorders statistics.
| Metric | Value | Source |
|---|---|---|
| Fry rate, native English bilinguals: English vs Spanish | 4.6% vs 2.9% (34 speakers: 22 women, 12 men) | Cantor-Cutiva et al., Journal of Voice 2023 |
| Fry rate, native Spanish bilinguals: English vs Spanish | 1.9% vs 0.9% | Cantor-Cutiva et al., Journal of Voice 2023 |
| Creaky syllables per speaker, Dutch men (n = 30) | Roughly 0-5% | van Hugte et al., Journal of Voice 2026 |
| Conversational entrainment (20 young women) | Significantly more fry with a high-fry partner | Borrie et al., Journal of Voice 2017 |
| Fry patterns in reading (5 women, 174 samples) | Single pulse 76%, period doubling 13%, other 11% | Proctor et al., Journal of Voice 2024 |
| Creak in breath groups, laryngeal dystonia vs controls (34 speakers) | 22% vs 5% | Marks et al., Journal of Voice 2024 |
| Automated creak, dystonia vs controls (60 Spanish speakers) | AUC 0.88 (vs glottic insufficiency 0.73) | Castro, Zubiaur Gomar and Marks, Journal of Voice 2026 |
| Unexpected creak as diagnostic cue (50 patients, 50 controls) | AUC .82; optimal cutoff 4.0% (sensitivity .78, specificity .84) | Fletcher et al., Journal of Speech, Language, and Hearing Research 2026 |
Context note: in 104 Finnish female students, creak correlated with strain (r = 0.40) but not with vocal fatigue symptoms (Laukkanen and Rantala, Folia Phoniatrica et Logopaedica 2021). Women who used fry in conversational reading cut it significantly when asked to speak clearly, and less so when asked to speak loudly (Behrman et al., Folia Phoniatrica et Logopaedica 2016), the kind of speaking-style adjustment covered in our voice coaching statistics. Machine classifiers separated high-creak from low-creak Finnish speech in 90 women with 71.1% accuracy (Laukkanen et al., Journal of Voice 2024).
Summary: Vocal Fry & Creaky Voice by the Numbers
| Metric | Value | Source |
|---|---|---|
| Young US women using fry (n = 34) | Approximately two-thirds | Wolk et al., Journal of Voice 2012 |
| Female undergraduates with at least one fry episode (n = 50) | 86% | Gibson, Journal of Voice 2017 |
| Self-reported fry, women vs men | 18% vs 6% | Cantor-Cutiva et al., Logopedics Phoniatrics Vocology 2018 |
| Studies measuring English creak prevalence | 10, covering 297 speakers | Dallaston and Docherty, PLOS ONE 2020 |
| Finnish women with ‘quite a lot’ of creak, 1990s to 2010s | 7% to 25.8% | Uusitalo et al., Journal of Voice 2024 |
| Finnish men with ‘quite a lot’ of creak, 1990s to 2010s | 5.9% to 18.1% | Uusitalo et al., Journal of Voice 2024 |
| Sydney teens over 30 years (28 vs 21) | Increase for girls, not boys | White, Penney and Cox, JASA Express Letters 2025 |
| NPR female broadcasters’ samples with brief fry | 22% of 267 | Stewart et al., Journal of Voice 2025 |
| Listeners in hiring-voice study | 800; normal voice preferred on all 5 judgments (p < .001) | Anderson et al., PLOS ONE 2014 |
| Fry effect on attribute ratings | Partial eta squared 0.393; male speakers unaffected | Taylor et al., Journal of Voice 2025 |
| Crowdsourced listeners on fry and intelligence | 463; effect depends on pitch and rate | Parker et al., Journal of Voice 2018 |
| Spontaneous phrase-final creak, 104 listeners | No overall hirability penalty | Bellavance and Levi, Language and Speech 2026 |
| Creak rating, older vs younger NZ listeners | 50.2 vs 53.5 | Calhoun and White, Language and Speech 2025 |
| Identification accuracy, fry vs modal voice | 85% vs 51% | Stewart, Kling and D’Agosto, Journal of Voice 2026 |
| Fry linked to ‘vain’ by graduate listeners | 56% | Ligon et al., Journal of Voice 2019 |
| Fry rate, native English bilinguals in English vs Spanish | 4.6% vs 2.9% | Cantor-Cutiva et al., Journal of Voice 2023 |
| Creaky syllables in Dutch men | Roughly 0-5% | van Hugte et al., Journal of Voice 2026 |
| Creak in breath groups, dystonia vs controls | 22% vs 5% | Marks et al., Journal of Voice 2024 |
| Automated creak diagnostic accuracy | AUC up to 0.88 | Castro, Zubiaur Gomar and Marks, Journal of Voice 2026 |
Methodology and Sources
Every figure above was read during research for this article in a peer-reviewed paper, its abstract (via the Europe PMC database) or its open-access full text. No survey or polling organization (Pew, Gallup or similar) has published data on vocal fry, so this roundup relies entirely on academic phonetics, sociolinguistics and voice science. Popular statistics that circulate without a primary paper were excluded.
- Wolk L, Abdelli-Beruh NB, Slavin D: Habitual use of vocal fry in young adult female speakers, Journal of Voice 2012
- Abdelli-Beruh NB et al.: Prevalence of vocal fry in young adult male American English speakers, Journal of Voice 2014 and Occurrence frequencies of acoustic patterns of vocal fry, Journal of Voice 2016
- Gibson TA: The role of lexical stress on the use of vocal fry, Journal of Voice 2017; Gibson TA et al.: Vocal fry use in adult female speakers exposed to two languages, Journal of Voice 2017
- Cantor-Cutiva LC et al.: Factors associated with vocal fry among college students, Logopedics Phoniatrics Vocology 2018 and The effect of bilingualism on production and perception of vocal fry, Journal of Voice 2023
- Dallaston K, Docherty G: The quantitative prevalence of creaky voice (vocal fry) in varieties of English: a systematic review, PLOS ONE 2020
- Uusitalo T et al.: Has the prevalence of creaky voice increased among Finnish university students from the 1990s to the 2010s?, Journal of Voice 2024
- White H, Penney J, Cox F: Change in the prevalence of creaky voice over time in Australian English, JASA Express Letters 2025
- Stewart CF et al.: Six decades of female professional radio broadcasters, Journal of Voice 2025; Stewart CF, Kling I, D’Agosto A: Modal register, vocal fry, and uptalk, Journal of Voice 2026
- Anderson RC, Klofstad CA, Mayew WJ, Venkatachalam M: Vocal fry may undermine the success of young women in the labor market, PLOS ONE 2014 (PMC full text)
- Taylor B et al.: Impact of vocal fry and speaker gender on listener perceptions, Journal of Voice 2025
- Parker MA et al.: Judgments of intelligence and likability of young adult female speakers, Journal of Voice 2018; Borrie SA et al.: Conversational entrainment of vocal fry, Journal of Voice 2017
- Venkatraman A et al.: Continuous vocal fry simulated in laboratory subjects, American Journal of Speech-Language Pathology 2018
- Bellavance SR, Levi SV: Reconsidering listener evaluations of creaky voice, Language and Speech 2026; Calhoun S, White H: The effect of age on affective meanings of uptalk and creak, Language and Speech 2025
- Ligon C et al.: Perceived desirability of vocal fry among female speech communication disorders graduate students, Journal of Voice 2019; Chappell W et al.: Social and stylistic correlates of vocal fry in a cappella performances, Journal of Voice 2020; Davidson L: Effects of pitch, gender, and prosodic context on the identification of creaky voice, Phonetica 2019
- van Hugte TBR et al.: Exploring interspeaker variation in creaky voice in Dutch, Journal of Voice 2026; Proctor K et al.: Vocal fry patterns while reading, Journal of Voice 2024; Behrman A et al.: The effect of loud voice and clear speech on the use of vocal fry in women, Folia Phoniatrica et Logopaedica 2016
- Marks KL et al.: The prevalence of creak across breath groups in adductor laryngeal dystonia, Journal of Voice 2024; Castro ME, Zubiaur Gomar FM, Marks K: Automated creak detection in Spanish speakers with and without AdLD, Journal of Voice 2026; Fletcher BN et al.: Linguistically informed automated estimates of creak in adductor laryngeal dystonia, JSLHR 2026
- Laukkanen AM, Rantala L: Relations between creaky voice and vocal symptoms of fatigue, Folia Phoniatrica et Logopaedica 2021; Laukkanen AM et al.: Can a machine distinguish high and low amount of social creak in speech?, Journal of Voice 2024
- Data watch: This is a thin-data topic. Sample sizes are small (often 20-50 speakers), most US prevalence studies sampled female college students, and studies use different detection methods (perceptual, acoustic, automated) and different formulas, so their percentages are not directly comparable. Most recent available data for US prevalence: Gibson 2017 and Cantor-Cutiva 2018; the Wolk (2012) and Anderson (2014) studies are more than ten years old. The Yuasa (2010) and Henton and Bladon (1988) figures were read only as summarized in the Dallaston and Docherty review, whose authors note that some values were extracted by eye from graphs; the original American Speech page for Yuasa returned HTTP 403. The full text of White, Penney and Cox (2025) was blocked (HTTP 403), so only its abstract-level findings are reported, without percentages. Anderson et al. present preference ratios in a figure, not a table, so exact per-attribute percentages are not quoted. Perception results conflict: Anderson (2014), Taylor (2025) and Venkatraman (2018) found penalties using deliberately produced fry, while Bellavance and Levi (2026) found none for spontaneous phrase-final creak. A 2026 Folia Phoniatrica et Logopaedica paper asking whether intense vocal fry is a ‘voice behaviour meme’ had no abstract available at the time of writing.
Last updated: October 3, 2026. We update this roundup quarterly, and the next refresh is expected when the full text of the 2025 Australian English creak trend study and the next Journal of Voice and Language and Speech perception studies on spontaneous creak become available.