How to expand your child's mean length of utterance at home

Last updated 2026-07-10

What mean length of utterance actually measures

MLU stands for mean length of utterance: the average number of morphemes in a child's spoken sentences. A morpheme is the smallest meaningful piece of language, so it can be a whole word like "dog" or a grammatical piece like the "-ing" in "running" or the "-s" in "dogs." Both count toward the total.

Speech-language pathologists rely on MLU because it's one of the most reliable, best-studied markers of early language growth. Roger Brown first described the stages in 1973, and the basic framework has held up across five decades of research ^[1]. A child saying "Dog bite" has an MLU of 2. A child saying "The dog is biting me" has an MLU of 7. That gap tells a clinician a lot. Why bother with this at home? Because you can measure it yourself, roughly, set a realistic target, and check whether what you're doing is actually working. Parents who know what they're aiming for tend to feel steadier than those waiting on a diagnosis or a therapy slot to open up.

Typical MLU by age

Brown's original research described five stages of language development, tied to MLU rather than age, since kids vary so much ^[1]. The American Speech-Language-Hearing Association (ASHA) and later researchers mapped typical MLU ranges to approximate ages ^[2].

These are averages, not cutoffs. A 36-month-old with an MLU of 2.2 sits behind typical development but may just be a late bloomer. A 48-month-old with an MLU below 2.5 almost certainly needs a speech therapy evaluation. Nobody should diagnose from a table alone. One number worth remembering: kids on a typical path gain roughly one morpheme per year in the toddler window, moving from around 1.0 at 12 to 18 months to around 4.0 by 48 months ^[1]. That's slow enough that home strategies can genuinely move things between appointments.

Calculating MLU yourself

You don't need language sample analysis software for a rough estimate that's still useful. Record 50 to 100 of your child's spontaneous utterances during natural play or a meal, write them down, and count the morphemes in each using these rules: every free-standing word counts as one; grammatical endings like "-s," "-ed," "-ing," "-er," and "-est" each add one more; irregular past tenses like "went" count as one morpheme total. Add everything up and divide by the number of utterances. So "She went to the store" comes to 5 morphemes (she / went / to / the / store), while "She goes to the stores" comes to 7 (she / go / -es / to / the / store / -s). A few cautions: don't count imitated speech, unintelligible utterances, or singing ^[1]. The sample needs to be conversational, not prompted. If you only catch 20 utterances in 30 minutes, that's still a usable baseline, just treat it as approximate. This won't replace a formal language sample from a speech-language pathologist, but it will tell you whether your strategies are moving things in the right direction week over week.

The "one level up" rule

The single most evidence-supported home technique for raising MLU is modeling one morpheme above your child's current level. Researchers call this "expansion," and ASHA describes it as a core strategy in naturalistic language intervention ^[2]. If your child says "Dog," you say "Dog runs." If they say "Dog runs," you say "The dog is running." You're not correcting them, just showing them the next rung on the ladder, once, naturally, inside a real exchange. The research backs this up. A 2011 meta-analysis in the American Journal of Speech-Language Pathology found that parent-implemented naturalistic language interventions, which use expansion as a core technique, produced significant gains in expressive language compared to no treatment ^[3]. Effect sizes ran modest to medium, and these were parent-implemented strategies, not intensive clinic-based therapy. Modest gains at home, stacked on top of clinic gains, add up. Keep it conversational, though. Expand everything your child says and it starts to feel like a test. Aim for expansion on maybe half of what they produce, and let the rest be a normal back-and-forth.

The routines that build MLU without trying too hard

Routines work because they're predictable. Your child already knows how bath time, breakfast, getting dressed, and a car ride go, and that familiarity frees up enough attention for language ^[4]. Bath time is one of the best windows there is. The vocabulary stays consistent (water, soap, pour, splash, wet, warm), the actions repeat, and your child is right in front of you with nothing else competing for attention. Say what's happening: "The water is going in." Pause. Wait. If they say "water in," expand it: "Yes, the water is going in the tub." Mealtimes are another strong window, especially if you narrate the food and actions without demanding responses. "I'm cutting the apple. Crunch. The apple is cold." This kind of narration, sometimes called self-talk or parallel talk, hands your child grammatically complete sentences during a moment they're already tuned in ^[2]. Car rides are underrated: there's nothing else to do, your child is captive, and you can run simple games like "I see a..." or comment on what you pass. Keep it low-pressure, since the goal isn't a quiz, it's exposure. Getting dressed and cleanup time work well for verbs specifically. "You're putting on your shoe. Now pull the sock up. Push your arm through." Verbs are often the missing piece holding a child's MLU down.

Expansion versus extension

These two techniques get confused constantly, but they work differently. Expansion repeats the child's utterance with the missing grammar filled in: child says "Dog eating," parent says "Yes, the dog is eating." Same meaning, more morphemes. Extension (sometimes called expatiation) adds new information instead: child says "Dog eating," parent says "The dog is eating his bone because he's hungry." You keep their meaning and layer more on top. For raw MLU growth, expansion tends to win in the short run because it spotlights the exact grammatical morphemes a child is missing. Extension does more for vocabulary and narrative skills, which matter later. Most speech-language pathologists weave both in rather than picking one. A study in the Journal of Speech and Hearing Research found that parents trained in these focused stimulation techniques saw significantly greater gains in their children's expressive language than untrained parents, with the effect showing up within 8 to 12 weeks of consistent use ^[5]. The catch is that "consistent" meant using the strategies across many interactions every day, not during one dedicated 10-minute practice session.

Which words and grammar targets move the needle fastest

Brown's original research identified 14 grammatical morphemes that children acquire in a strikingly consistent order ^[1]. Targeting the ones just ahead of where your child sits raises MLU faster than targeting random vocabulary. The earliest morphemes to appear, and the first worth targeting, are:

• Present progressive "-ing" (dog running)
• Plural "-s" (dogs, cats)
• Possessive "-'s" (daddy's shoe)
• Regular past tense "-ed" (he walked)
• Articles "a" and "the"
• Auxiliary "is" (he is going)

A child at MLU 1.5 probably isn't using any of these consistently yet. Start with "-ing" endings and plural "-s," since they appear earliest and attach to high-frequency words your child already knows. Verbs are the other big lever. Children with low MLU often have a verb vocabulary of fewer than 10 words, and you can't build multi-word sentences without verbs. Targeting action words during play (push, pull, eat, jump, go, open, fall, throw) produces faster MLU gains than adding more nouns, because verbs demand and invite word combinations ^[6][11]. The same logic applies if your child uses AAC: adding grammatical morpheme buttons and modeling their use during routines runs parallel to spoken expansion, and the overview of aac devices covers that in more detail.

How does play fit into raising MLU?

Play isn't a break from language work, it's the actual work. Language doesn't grow because you scheduled twenty minutes for it. It grows when a child is engaged and motivated, in a low-pressure exchange with someone following their lead.

Joint attention is what makes this work. When a child and caregiver both focus on the same thing, language input is far more likely to stick ^[4]. So follow your child's interest instead of your plan. If they're pushing a truck back and forth, don't drag them over to the farm set. Get on the floor and narrate the truck: "The truck goes fast. It crashed! The truck fell down."

One technique worth knowing is sabotage play, which has research behind it. Put a favorite toy slightly out of reach, hand them a container they can't open, or leave out a key piece of a toy set. That mild frustration creates communication pressure in a safe setting, and the child has to request something. When they do, even without words, model the language back: "You want the ball? Say 'ball' or 'want ball'."

Wait time is the part most parents rush past. After a communication opportunity, pause for a full 5 seconds. It feels endless, but that silence isn't awkward, it's an invitation. Many children need 4 to 6 seconds to put a response together, especially with any processing differences ^[2].

For kids with autism, social motivation around play-based conversation can look different. The strategies still work, but they often need to sit more heavily inside the child's specific interests. The page on autism spectrum speech therapy goes further into adapting these approaches.

What should parents avoid when trying to raise MLU?

The most common mistake is asking too many questions. Questions pressure a child to perform and can actually cut down how much language they produce ^[2]. A parent spending a whole meal on "What's that? What color is it? What's in your bowl?" is running drills, not talking. Shift toward comments instead: "That's a big piece of broccoli. It's green and bumpy."

Correction backfires, too. If your child says "I goed to the park" and you answer "No, you mean went," they hear criticism, not instruction. Expansion does the same job without the sting: "Oh, you went to the park! That sounds fun." The right form shows up naturally in your reply, no correction needed.

Watch the complexity of your own language as well. If your child is at MLU 2.0 and you're modeling 8-word sentences, you've gotten too far ahead. Input needs to sit one level up, not five.

Rushing through routines dulls the whole effort. Plenty of parents narrate bath time while mentally planning dinner, but kids feel the difference between a distracted comment and a real exchange. Ten focused minutes beats forty distracted ones.

The costliest mistake, though, is waiting it out. The research on early intervention is consistent: earlier support produces better outcomes, and sitting past age 3 with no professional input is a real cost when the delay is significant ^[7][10].

How long does it take to see MLU gains from home strategies?

Honestly, it varies a lot, and nobody has clean population-level data on home-only timelines because most studies fold in some professional coaching alongside the home work.

The closest evidence comes from parent-implemented naturalistic intervention studies. Several show measurable expressive language gains within 8 to 16 weeks of consistent daily use ^[3][5]. "Measurable" here means statistically significant against control groups, not necessarily dramatic in daily life.

For a child at MLU 1.5 in a responsive communication environment every day, a gain of 0.5 to 1.0 morphemes over 3 to 4 months is realistic from home strategies alone. That's meaningful: it can move a child from single words to two-word combinations, or from two-word combinations to basic three-word sentences.

Children with additional diagnoses, such as autism, apraxia of speech, or hearing differences, generally need more intensive, specialized support alongside the home work. The home work still counts for a lot, but on its own it's probably not enough.

Track progress monthly. Collect a rough sample (even just noting utterances on your phone during a car ride), count the morphemes, and compare month to month. No movement after 2 to 3 months of daily effort is a clear signal to add professional support rather than keep waiting.

When should you involve a speech-language pathologist?

Home strategies are worth doing on their own, and they work even better with a speech-language pathologist also in the picture. Those two things aren't in competition.

ASHA recommends evaluation if a child isn't using 50 words by 24 months, isn't combining two words by 24 months, or has an MLU well below age expectations ^[2]. The American Academy of Pediatrics (AAP) recommends that pediatricians screen for language delays at the 9-, 18-, and 30-month well-child visits, and refer for evaluation whenever a screen flags a delay ^[7].

If your child's MLU is more than 6 months behind the typical range for their age and you haven't seen an SLP yet, make the call. Most states run publicly funded early intervention programs for children under 36 months that provide services at no cost to families, under the Individuals with Disabilities Education Act (IDEA) Part C ^[8].

For children 3 and up, services shift to Part B of IDEA, run through school districts. Eligibility rules vary by state, but a language delay that affects educational performance typically qualifies ^[8].

If you want support between appointments, or waitlists in your area run long (they often do), tools like online speech therapy or AI-assisted practice apps can supplement clinic work. Little Words was built specifically for neurodivergent kids in that in-between space; a short quiz at littlewords.ai/start will tell you whether it fits your child's profile. And if your child shows repetitive or scripted speech alongside language delays, the piece on echolalia covers how to tell functional echolalia from non-functional, and how it connects to MLU development.

What does the research say about parent-implemented language intervention?

The evidence here is genuinely encouraging, with a few caveats worth knowing.

A 2011 meta-analysis in the American Journal of Speech-Language Pathology by Roberts and Kaiser looked at 18 studies of parent-implemented language interventions. It found consistent, positive effects on children's expressive vocabulary and MLU, with the strongest effects in children under 36 months ^[3]. Effect sizes ran small to moderate: in practical terms, home intervention moves outcomes meaningfully but doesn't replace specialist input for children with significant delays.

The ingredient that mattered most across successful studies was parent responsiveness: how much parents followed the child's lead, backed off directive questioning, and consistently expanded on what the child said ^[6]. Responsiveness predicted child outcomes better than any specific technique did.

Quality of input beats quantity, too. A 2017 study in the American Journal of Speech-Language Pathology found that the density of child-directed speech built on back-and-forth conversational turns, not the raw volume of words heard, was the strongest predictor of language growth ^[9]. That reframes "talk more" into something sharper: talk responsively, more than frequently.

Parents who get even brief coaching (one to four sessions with an SLP on these techniques) implement them with noticeably better fidelity than parents working from written strategies alone ^[5]. Where professional access is thin, videos of real parent-child interaction reviewed by a teletherapy SLP make a workable middle ground.

Common questions about MLU

At 24 months, a typical MLU falls between 1.5 and 2.0 morphemes, which usually looks like single words or simple two-word combos such as "more milk" or "daddy go." If your 2-year-old is consistently below MLU 1.5, or still only using single words with no combinations at all, that's worth a speech-language evaluation.

You can absolutely raise your child's MLU without a therapist in the room. Expanding on what your child says, following their lead, and weaving language into everyday routines all have research behind them. That said, kids with significant delays or additional diagnoses like autism or apraxia usually need professional support too. It's not really a choice between home strategies and therapy; the two work best together.

There's no magic number of practice sessions. Research points to consistency across the day mattering more than one long, dedicated session. Try working expansion into at least three natural routines daily, maybe a meal, playtime, and something like bath or dressing. Short, frequent, responsive exchanges woven into your normal day beat occasional intense drilling.

Reading together does help, mainly because shared books build vocabulary and narrative skills and create joint attention, all of which feed into MLU growth. The trick is reading interactively: pause, point, comment, let your child jump in. Firing off "What's that?" again and again turns a book into a quiz. Saying something like "Look, the dog is jumping" models the grammar you actually want them absorbing.

Imitated and echolalic speech doesn't count toward a formal MLU score. But functional echolalia, when a child reuses a memorized phrase to communicate something real, gets treated differently by many clinicians. If most of your child's speech is echolalic, the echolalia piece on our site walks through that distinction and which strategies fit, and an SLP familiar with these patterns should really be guiding the approach.

MLU isn't quite the same as counting words in a sentence. It counts morphemes: "dogs" is one word but two morphemes (dog plus the plural -s), so a five-word sentence might carry seven morphemes once you count endings. Word counting is simpler and gets you close enough for tracking progress at home; a formal analysis counts morphemes instead.

Screen time doesn't do much for MLU compared to live, responsive conversation, especially under age 3, per American Academy of Pediatrics guidance. Older kids can pick up vocabulary from video, but grammatical development still depends on back-and-forth interaction. If your child does watch something, watching together and talking about it does more than letting them watch alone.

A 4-year-old still at MLU 2 is meaningfully behind: typical MLU at 48 months runs 4.0 to 4.5, so an MLU of 2.0 represents roughly a two-year gap. That calls for evaluation by a speech-language pathologist, and possibly assessment for childhood apraxia of speech, autism, or hearing loss. The evidence strongly favors getting support early rather than waiting.

Expressive vocabulary and MLU measure different things: vocabulary is how many different words a child uses, while MLU measures how they combine and inflect those words. Under 24 months, a child generally needs around 50 words before word combinations reliably show up. So if your child has fewer than 50 words, focus on vocabulary first; past that point, shift toward expanding MLU.

MLU measures language production, not general intelligence. Plenty of children with typical or high intelligence show delayed MLU because of motor speech differences like apraxia, hearing loss, or autism-related language processing. It's a language metric, not a cognitive one, so try not to read a child's language output as a stand-in for their intellectual potential.

Bilingual exposure can lower MLU scores in either language taken alone, since bilingual children are splitting their language across two systems. A fair assessment should look at MLU in both languages or use bilingual norms, so let your SLP know if your child hears more than one language at home. It changes how the numbers should be interpreted.

Most consumer speech apps don't have much research behind them specifically for MLU, if we're being honest. What lines up with the evidence is anything that builds in responsive back-and-forth, models grammatically expanded language, and follows the child's lead. Little Words was built around these principles for neurodivergent kids. Beyond apps, the things that actually move the needle are a parent who knows what to model, steady routines, and access to an SLP even now and then.

Sources

1. Brown, R. (1973). A First Language: The Early Stages. Harvard University Press.: Brown's five-stage MLU framework and the acquisition order of 14 grammatical morphemes; children gain roughly one morpheme per year in the toddler window.
2. ASHA – Late Language Emergence Practice Portal: ASHA identifies expansion modeling, reduced directive questioning, and wait time as core naturalistic language intervention strategies; recommends evaluation if a child is not combining two words by 24 months.
3. Roberts, M.Y., & Kaiser, A.P. (2011). The effectiveness of parent-implemented language interventions: A meta-analysis. American Journal of Speech-Language Pathology, 20(3), 180–199.: Meta-analysis of 18 studies found parent-implemented naturalistic language interventions produced significant gains in children's expressive language and MLU, with strongest effects in children under 36 months.
4. National Institute on Deafness and Other Communication Disorders (NIDCD) – Speech and Language Developmental Milestones: Joint attention between caregiver and child is a core mechanism through which language input becomes language learning.
5. Girolametto, L., Pearce, P.S., & Weitzman, E. (1996). Interactive focused stimulation for toddlers with expressive vocabulary delays. Journal of Speech and Hearing Research, 39(6), 1274–1283.: Parents trained in focused stimulation (expansion and extension) techniques saw significant expressive language gains in their children within 8 to 12 weeks, with training fidelity as a key predictor.
6. Tamis-LeMonda, C.S., Bornstein, M.H., & Baumwell, L. (2001). Maternal responsiveness and children's achievement of language milestones. Child Development, 72(3), 748–767.: Parent responsiveness (following child's lead, expanding utterances) was more predictive of language milestone achievement than the specific technique used.
7. American Academy of Pediatrics – Developmental Surveillance and Screening: AAP recommends pediatricians screen for language delays at 9-, 18-, and 30-month well-child visits and refer for evaluation when screening suggests a delay.
8. U.S. Department of Education – IDEA (Individuals with Disabilities Education Act): IDEA Part C funds publicly available early intervention services at no cost to families for children under 36 months with developmental delays; Part B provides services through school districts for children 3 and older.
9. Gilkerson, J., Richards, J.A., Warren, S.F., et al. (2017). Mapping the early language environment using all-day recordings and automated analysis. American Journal of Speech-Language Pathology, 26(2), 248–265.: Density of conversational turns in child-directed speech was the strongest predictor of language growth, stronger than total word count heard.
10. Rescorla, L. (2002). Language and reading outcomes to age 9 in late-talking toddlers. Journal of Speech, Language, and Hearing Research, 45(2), 360–371.: Late talkers who did not catch up by age 3 showed persistent language differences through age 9, supporting the case for early rather than delayed intervention.
11. Fey, M.E., Cleave, P.L., Long, S.H., & Hughes, D.L. (1993). Two approaches to the facilitation of grammar in children with language impairment. Journal of Speech and Hearing Research, 36(1), 141–157.: Targeting grammatical morphemes in Brown's acquisition order (beginning with present progressive -ing and plural -s) produced faster MLU gains than random vocabulary targeting.