Speech Sound Age of Acquisition: When Children Master Each Sound

When each speech sound is acquired — the short answer

Speech sounds are not learned all at once — they arrive in a remarkably consistent order. The most rigorous answer comes from Crowe & McLeod’s 2020 review, which pooled 15 studies of 18,907 US children and reported, for every English consonant, the age band in which 90% of children say it correctly. At that criterion: /b, n, m, p, h, w, d/ are acquired by 2;0–2;11; /g, k, f, t, ng, y/ by 3;0–3;11; /v, j, s, ch, l, sh, z/ by 4;0–4;11; /r/, voiced /th/, and /zh/ by 5;0–5;11; and voiceless /th/ by 6;0–6;11.¹

The notation “5;0–5;11” means years;months — so R being acquired at 5;0–5;11 means the 90% mark is reached at some point between the fifth and sixth birthday. In plain terms: a 4-year-old still working on R is on schedule, and most children have it before they turn 6.¹

The speech sound age-of-acquisition table (90% criterion)

Here is the complete consonant timetable. Each row gives the age band in which 90% of US children in the pooled studies produced the sound correctly. Every sound in the two latest rows has its own SpeechStep guide and practice page — R, S, Z, TH, L, SH, and CH — where your child can record the sound and get instant feedback.¹

The same timetable by sound class

The order also holds at the level of whole sound classes. In Crowe & McLeod’s data, on average all plosives, nasals, and glides were acquired by 3;11; all affricates by 4;11; all liquids by 5;11; and all fricatives by 6;11.¹

Why published charts disagree

If you have seen other charts with different ages, you are not imagining it — and the reason has been understood for over 50 years. Sander’s classic 1972 paper drew the key distinction: the “customary” age of production, when half of children say a sound correctly, versus mastery, when 90% do. A chart built on the 50% criterion will always show younger ages than one built on the 90% criterion, even from the same data.³

On top of the criterion, the underlying studies themselves differ. Crowe & McLeod note that variation was evident across studies, resulting from different assessments, criteria, and cohorts. That is exactly why this page uses one source, one criterion, and prints the criterion in the table caption — so you know precisely what each age means.¹

A second source of confusion: ASHA’s current consumer chart is deliberately short. It lists “By 3 years — says m, n, h, w, p, b, t, d, k, g, and f in words. By 4 years — says y and v in words,” and summarizes the rest as “Most children can say almost all speech sounds correctly by 4 years old.” It no longer gives ages for S, R, TH, L, SH, CH, or Z — the exact sounds parents ask about most. This table fills that gap with the peer-reviewed norms.⁴

The latest-developing speech sounds: S, Z, L, SH, CH, R, and TH

Everything after age 4 belongs to a small, famous group. In the 4;0–4;11 band sit /v, j, s, ch, l, sh, z/; in the 5;0–5;11 band, /r/, voiced /th/, and /zh/; and voiceless /th/ arrives last, at 6;0–6;11. So a 4- or 5-year-old still polishing R or TH is inside the typical window for that age band.^1,7

The /s/ and /z/ row also answers the most common lisp question. ASHA does not publish a lisp “age of elimination” on its consumer chart — the defensible anchor is the norm itself: /s/ and /z/ are acquired at 4;0–4;11 at the 90% criterion. An S that still sounds like “th” past that band is worth assessing; our frontal lisp and lisp in children guides cover what a lisp is and what to do.^1,4

Do these ages hold across languages and for bilingual children?

The table above is built from US English data, but the overall timetable is close to universal. McLeod & Crowe’s companion review pooled 26,007 children across 27 languages and found that most of the world’s consonants were acquired by age 5;0 — by which point children produced at least 93% of consonants correctly.²

For bilingual families, that is the practical takeaway: whatever the language mix, “most consonants in place around age 5” is the cross-linguistic norm. Individual sounds can differ — a sound that is late in English may be early in another language and vice versa — so use the English table for English targets, and the age-5 benchmark for the big picture.²

When a sound is late — and when to act

Sounds that stay wrong well past their band do not always fix themselves. Flipsen’s review of persistent and residual speech errors found that 1%–2% of young adults still exhibit residual or persistent speech errors — chiefly distortions of /s/ and /r/, the same late sounds in the bottom rows of the table. That is the honest case against indefinitely waiting on a stubborn S or R.⁵

ASHA’s Practice Portal describes how clinicians assess and treat speech sound disorders, and you can find a certified SLP through ASHA’s ProFind directory. SpeechStep is not a diagnosis and not a replacement for an SLP — it is the daily, structured practice between sessions (or before them), with AI feedback on each attempt and clear guidance for parents.^8,10

Turn the table into practice

The table tells you which sound is due; practice is what moves it. Typical development is a sequence, not a lottery — which is why short, frequent practice on the one sound that is due works better than drilling everything at once.⁹

• Your child is 4–5 and S, Z, L, SH, or CH is shaky → start with that sound’s guide and practice page.
• Your child is 5–6 and R or TH is not in yet → the R Sound and TH Sound guides show what the sound should look like and let your child record and practice it.
• Not sure which sound to look at → run the free milestone checker below for a two-minute, age-matched starting point.

Each practice page plays a model of the sound, records your child’s attempt, and gives instant, encouraging feedback — a few minutes a day, at home. And if what you hear worries you, the checker’s results make a concrete starting point for a conversation with an SLP.

Frequently asked questions

References

10 sources from authoritative bodies. Last reviewed August 2026.

1. 1.Peer-reviewedCrowe & McLeod — Children’s English Consonant Acquisition in the United States: A Review — Systematic review (AJSLP), 2020.
2. 2.Peer-reviewedMcLeod & Crowe — Children’s Consonant Acquisition in 27 Languages: A Cross-Linguistic Review — Cross-linguistic systematic review (AJSLP), 2018.
3. 3.Peer-reviewedSander — When Are Speech Sounds Learned? — Classic primary study (JSHD), 1972.
4. 4.ASHASpeech Sound Disorders — Consumer page + sound chart.
5. 5.Peer-reviewedFlipsen — Emergence and Prevalence of Persistent and Residual Speech Errors — Review (Seminars in Speech and Language), 2015.
6. 6.NIDCDSpeech and Language Developmental Milestones — Fact sheet.
7. 7.ASHACommunication Milestones: 4 to 5 Years — Developmental milestones.
8. 8.ASHASpeech Sound Disorders: Articulation and Phonology — Practice Portal page.
9. 9.ASHATypical Speech and Language Development — Consumer page.
10. 10.ASHAASHA ProFind — Find a Certified SLP — Clinician directory.