At-home speech therapy: exercises, routines, and what actually works

Last updated 2026-07-09

You can absolutely run speech practice at home, and doing so doesn't compete with a licensed speech-language pathologist (SLP), it's what makes the professional sessions actually stick. ASHA notes that children who get parent-implemented practice between sessions make faster progress than kids who only show up for weekly appointments ^[1]. That tracks: a 45-minute clinic visit once a week is 45 minutes. A parent doing 15 minutes a day is 105 minutes a week. Volume matters. Plenty of parents worry they'll get it wrong, and that worry is understandable, but it shouldn't stop you. Research on parent-implemented language intervention, including the Hanen Program evidence base, shows coached parents produce real gains in their children's vocabulary and sentence length ^[2]. You don't need a graduate degree, just consistency, the right targets, and a feedback loop with your child's SLP. If your child doesn't have one yet, you can learn how to get an evaluation, and if they're under three, early intervention services are free under federal law and include home-based therapy.

Building a routine that actually holds

Structure matters more than duration. A chaotic 30-minute session helps less than a focused 10-minute one. Start by picking one or two targets at a time, ideally set by your SLP. Without one right now, focus on sounds or words your child is already close to producing: targeting sounds already in their inventory, ones they can make at all even inconsistently, is more efficient than jumping to sounds they can't approximate yet ^[3].

Fold the practice into routines you already have. Bathtime, meals, and car rides work well because repetition happens on its own, and you're not carving out a separate "speech time" your child dreads. Keep a simple log, too, a notes app is fine, of what you practiced, what your child did, and anything that surprised you. Sharing that with your SLP lets them adjust targets without relying only on what they see in the room. And stop while your child is still engaged. Ending on a high note keeps motivation intact for next time: five successful trials beat fifteen where the last ten were refusals.

That adds up to 12-20 minutes total, which is sustainable every day.

Exercises that work for toddlers and late talkers

Late talkers, roughly defined as children with fewer than 50 words or no word combinations by age 24 months, respond well to a specific cluster of techniques with the strongest evidence available for home use ^[4].

Model without demanding: say the target word naturally during play, without asking your child to repeat it. "Ball. Ball. Roll the ball." You're flooding the input, not running a drill. This approach, called recasting, has consistent support in the literature ^[5]. Parallel talk works alongside it: narrate what your child is doing in simple language ("You're pouring. Pouring the water. It's wet."), keeping your utterances one word longer than what they say now. No words yet? Use single words. One-word utterances? Move to two-word phrases.

Try expectant pausing too: set up a situation where your child needs something, then genuinely wait. Hold the toy out, look at them, stay quiet for five to ten seconds. Most parents fill the silence immediately; don't. The pause creates communicative pressure without the pressure of a direct demand. For children not yet combining sounds into words, sound play helps build the motor habit of intentional vocalization: blow raspberries, imitate each other's noises, make animal sounds, before you worry about meaning.

When you share books, skip "what's that?" and comment instead: "Dog. Big dog. He's running." Questions interrupt the flow of input; comments add to it. The difference sounds small, but the effect size in research isn't ^[5]. For children with suspected apraxia of speech, the approach shifts toward motor-based practice with precise feedback, and childhood apraxia of speech covers what that looks like at home.

Choosing the right sounds, in the right order

Speech sounds don't all develop at the same age, and expecting a two-year-old to produce a clean "r" sets you up to drill the wrong thing entirely. ASHA publishes developmental norms for speech sound acquisition ^[3], roughly in these tiers:

Practice sounds that are developmentally expected for your child's age or slightly ahead. A four-year-old missing "k" and "g" is a reasonable target; a five-year-old missing "r" is within normal range, and drilling it at home may just frustrate everyone for no gain.

Once you pick a sound, move through word positions in order: initial ("cat"), medial ("baking"), then final ("back"). Most children find initial position easiest. Once they've got it consistently in single words, move to phrases, then short sentences, then conversation. That progression is called generalization, and it's where most home practice stalls, so don't skip it.

What works for kids with autism

Autistic children often have communication profiles that don't fit the typical late-talker mold. Some are minimally verbal. Some have large vocabularies but struggle with the social use of language. Some use echolalia as a functional communication strategy, so the exercises need to match the profile in front of you rather than a generic script.

For minimally verbal children, the research support is strongest for Naturalistic Developmental Behavioral Interventions (NDBIs), which include approaches like JASPER (Joint Attention, Symbolic Play, Engagement, and Regulation) and the Early Start Denver Model. Both are teachable to parents and have randomized trial evidence behind them ^[6]. The core idea is to follow your child's lead into activities they're already engaged with, then build communicative openings inside those activities.

For children who use echolalia, the goal isn't to stop it. It's often meaningful and functional, and echolalia meaning covers how to read what your child is communicating through it and build on it rather than suppress it. For children who use AAC devices, home practice with their system matters a great deal: meaningful AAC use requires dense, consistent modeling by communication partners, and thirty exposures to a symbol in a single day isn't too many ^[7]. This guide covers AAC and related strategies in more depth.

One principle holds across every autism communication profile: lower the demand load during practice. High-demand environments raise anxiety and cut communication, while low-demand, child-led activities produce more spontaneous language, which is the whole point.

Do the apps actually help?

It depends entirely on which app and how you use it. Some are grounded in real speech-language principles; others are word-labeling games with a therapy label slapped on. Articulation Station (by Little Bee Speech) organizes practice by sound and word position, and Speechy Mushy is SLP-designed. Both help parents stay organized and give children a low-stakes way to hear target words repeatedly, which works fine as a drill supplement.

Other apps marketed as speech therapy are really just picture-naming games. Pointing at pictures and hearing their names builds receptive vocabulary in typically developing toddlers, but for children with significant delays, that passive exposure isn't enough on its own. The American Academy of Pediatrics recommends avoiding screen time for children under 18 months (outside of video chatting), and that children 18-24 months only use screens with a caregiver present ^[8]. For speech practice, the interactive part is the whole point: an app your child uses alone in a corner isn't the same as one you use together, with narration, pausing, and conversation woven around it.

Little Words is built as an AI speech companion for neurodivergent kids, meant to sit alongside therapy rather than replace it. If you want a structured way to run daily practice without winging it, the start quiz can help identify which exercises fit your child's current communication level.

Worth saying plainly: no app has randomized controlled trial evidence showing it produces the same outcomes as SLP-led therapy. Apps are organizational tools and practice scaffolds, not clinicians, and this article is general information rather than a stand-in for advice from your child's own speech-language pathologist.

How much home practice actually helps depends on your child's age and diagnosis, but the research gives some real numbers to work with, and the short version is that showing up daily matters more than long sessions.

A 2018 Cochrane review of early language interventions found that frequency and intensity of practice correlated positively with outcomes: parent-implemented programs averaging 8 to 12 weeks of daily or near-daily practice showed stronger effects than lower-intensity ones^[9]. That's not permission to drill your child into exhaustion. It just means consistency beats occasional marathon sessions. Most SLPs recommend 10 to 20 minutes of focused practice a day for kids in therapy, though children under 3 may only hold attention for 5 to 8 minutes, and that's fine. Two short sessions beat one long one your child abandons halfway through. If your child has childhood apraxia of speech, the bar is higher: Apraxia Kids and current motor learning research point to 60 to 100 trials per session with frequent feedback^[10]. A professional needs to set the targets first, but the actual practice can happen at home. Build in breaks, too. A child who starts associating talking with pressure will just avoid talking, which defeats the whole point.

You don't need much in the way of materials, honestly. Blocks, toy cars, play food, a ball, books, whatever your child is currently obsessed with, all of it works fine as a vehicle for language practice. If you want a small kit anyway, a few things are genuinely worth having. Articulation flashcards organized by sound (cheap on Teachers Pay Teachers, or free from SLP blogs) are useful for structured trials, though not for endless drilling. A small mirror helps too, since a child can't see their own tongue but can compare their reflection to yours when you're working on sound placement. A reward chart works well if you reward trying rather than getting it right, since early on the goal is participation, not accuracy. And a notebook where you jot down what you tried and what your child said spontaneously turns out to be useful during SLP check-ins. Beyond that, skip the expensive therapy toys and app subscriptions. What actually matters is attention, time, and clear targets, and the SLP supplies the targets.

Home practice supports professional care, it doesn't stand in for it. The American Academy of Pediatrics and ASHA both recommend evaluation if a child^[1][4]:

• Has no words by 12 months
• Has fewer than 6 words by 18 months
• Has fewer than 50 words or no two-word combinations by 24 months
• Loses language skills at any age
• Is very difficult to understand by age 3 (less than 75% intelligibility to familiar adults)
• Has a feeding or swallowing issue alongside the speech concern

These aren't gentle suggestions. Early intervention under IDEA Part C is federally mandated and free for children birth to age 3^[11], and school districts must evaluate school-age children at no cost when a disability is suspected to affect their education. If in-person access is limited, online speech therapy is a legitimate option too: telehealth SLP services have shown outcomes comparable to in-person therapy for most speech and language goals^[12]. But no amount of home exercise replaces an evaluation. If something feels off, get it checked. Finding out your child is developing typically costs you nothing. Waiting when something's actually wrong costs you time you can't get back.

The evidence for parent-implemented therapy is stronger than a lot of parents, and some clinicians, give it credit for. Roberts and Kaiser's 2011 meta-analysis in the American Journal of Speech-Language Pathology looked at 18 studies of parent-implemented interventions and found significant effects on both expressive and receptive language, concluding that "parent-implemented language interventions are effective for improving child language outcomes"^[2]. The Hanen Program, which trains parents in responsive interaction, has peer-reviewed studies behind it showing gains in vocabulary, sentence length, and spontaneous communication^[2]. For autism specifically, the Early Start Denver Model, which leans heavily on parent coaching, produced gains in language and adaptive behavior in a 2010 randomized controlled trial in Pediatrics, with the parent-coached group outperforming community-treatment controls^[6].

Nobody has precise numbers on how many weekly minutes of home practice produce gains independent of clinic time. What the research does suggest is that how responsive a parent is, and the quality of their language modeling, matters more than raw minutes logged^[5]. Ten minutes a day from an engaged parent likely beats thirty minutes of going through the motions. Worth knowing, though: most of this research involved parents who'd already been coached by an SLP. Home practice without that guidance, where parents are guessing at targets, has weaker evidence behind it. The two work together. Either one alone does less than both combined.

There are decent free resources out there too. ASHA's public website has reliable, if basic, information on milestones and red flags^[1]. The Hanen Centre (hanen.org) publishes free articles and tip sheets that explain their techniques clearly, even though the full programs require professional training. SuperDuper Publications has free worksheets by speech goal, and Teachers Pay Teachers has thousands of low-cost or free articulation and language materials. Early intervention under IDEA is free for qualifying children under 3, including home visits from SLPs^[11], and school-based services are free for ages 3 to 21 under IDEA Part B. For AAC, PrAACtical AAC (praacticalaac.org) has free content on modeling AAC at home and printable communication boards.

The one thing that costs money is a licensed SLP's time, and for some kids that expense is necessary. Free resources are a foundation, not a full solution, especially for moderate to severe delays. If you're trying to figure out where your child stands and what to actually do at home, Little Words (littlewords.ai/start) has a free quiz that can help point you toward next steps.

Frequently asked questions

Can home practice replace working with a speech-language pathologist?

No. Practicing at home works best alongside SLP-led therapy, not instead of it. ASHA's guidance is clear that parent-implemented strategies produce better outcomes when parents get coaching from a licensed professional first. If a child has a significant delay, feeding trouble, or a suspected diagnosis like apraxia or autism, a professional evaluation matters. What you do at home reinforces what therapy targets, but it doesn't substitute for clinical assessment and planning.

What helps a 2-year-old who isn't talking yet?

Lean on input-heavy strategies rather than drilling: narrate what your child is doing in short one-to-two word phrases, pause after setting up a chance for them to communicate, and imitate their sounds so they see that communication runs both ways. Don't push for repetition. If your child genuinely has no words at 24 months, that meets the threshold for an early intervention referral under IDEA, which is free for children under 3.

How long before you see results?

Most research on parent-implemented language work reports measurable gains after 8 to 12 weeks of consistent daily practice. Results vary a lot depending on the child's diagnosis, how accurate the home targets are, and how much professional support the parent has had. Parents often notice new word attempts within two to four weeks of steady modeling, though articulation accuracy usually takes longer, especially for sounds that need precise motor coordination.

Is there a best app for toddlers?

No single app has randomized controlled trial evidence showing it beats the others for toddlers with delays. Among the options built by SLPs, Articulation Station (Little Bee Speech) is widely used by therapists for sound practice. For toddlers with language delays, look for apps that prompt interaction with a caregiver rather than solo play. The AAP recommends screen use for under-2s only with a caregiver present.

Are exercises different for kids with autism?

Yes, quite a bit. Children with autism often do better with naturalistic, low-demand approaches than structured drills. Naturalistic Developmental Behavioral Interventions like JASPER and the Early Start Denver Model can be taught to parents and have randomized trial evidence behind them. Children who use AAC need dense symbol modeling from caregivers throughout the day, and for minimally verbal children, the goal is building any intentional communication, not necessarily spoken words.

What helps with articulation at home?

Articulation practice works best as word-level drills that move through initial, medial, and final positions, then build up to phrases and sentences. A mirror helps your child see tongue and lip placement. Stick to sounds that are developmentally expected for your child's age, using ASHA's norms as a guide, and aim for 10 to 20 trials a session. Feedback that's immediate and specific ("you said the 't' at the end, great") works better than generic praise.

How do I know which sounds to target?

Ask your child's SLP for specific targets. If you don't have one yet, ASHA's developmental sound norms are a reasonable guide: p, b, m, h, and w should be present by age 3; k, g, and f by age 4; l by age 6; r and s by ages 7 to 8. Aim for sounds that are expected for your child's age or just slightly ahead. Drilling sounds that aren't developmentally expected yet just wastes practice time and frustrates everyone.

Can these exercises help with feeding and swallowing?

Some exercises overlap, since oral motor work involves lip and tongue movement either way, but feeding and swallowing therapy is its own specialty. Oral motor exercises done without guidance for a child with a swallowing issue can carry real risk. If feeding difficulties show up alongside speech concerns, get an evaluation from an SLP with feeding specialization or a feeding clinic rather than using generic speech exercises for suspected dysphagia.

What's the difference between a speech delay and a language delay?

Speech delay is about difficulty producing sounds clearly and accurately. Language delay is about difficulty understanding or using words, sentences, and communication more broadly. Plenty of children have both. A child with a speech delay might have rich language but poor intelligibility; a child with a language delay might have clear articulation but a very limited vocabulary or trouble following directions. An SLP evaluation sorts out which one you're dealing with, because the treatment differs.

How do I keep a toddler engaged during practice?

Follow the child's lead and practice inside whatever they're already doing, rather than pulling them into a separate "speech lesson." Keep sessions short, stop before they check out, and rely on physical play and natural rewards rather than sticker charts for young children. If blocks have stopped being interesting, switch to water play, playdough, or something outdoors. A child who's genuinely engaged gives you better trials than one complying under protest.

Is it okay to start at home before seeing an SLP?

For mild concerns, using parent-implemented strategies while you wait for an evaluation is unlikely to cause harm and may help, as long as you're sticking to developmentally appropriate techniques like modeling, parallel talk, and expectant pausing rather than intense drilling. For suspected apraxia, stuttering, or autism, though, unsupported home practice risks targeting the wrong things. The American Academy of Pediatrics recommends getting an evaluation whenever a developmental concern comes up, rather than waiting it out.

What does "modeling language" actually mean?

It means saying target words and phrases naturally in context without asking the child to repeat them. During play, you name objects, narrate what's happening, and comment on the action. The goal is high-quality input pitched slightly above what the child can currently say. Research shows children whose caregivers give dense, responsive language models build vocabulary faster. You're not quizzing them. You're just filling the environment with the language you want them to pick up.

Where can parents find free resources?

ASHA's consumer site has free milestone checklists and strategy tips, and Hanen.org offers free parent articles drawn from their evidence-based programs. For children under 3, early intervention evaluations and services are free under federal IDEA law, and school districts must evaluate school-age children at no cost if a disability affecting learning is suspected. PrAACtical AAC (praacticalaac.org) has free resources for families using AAC systems.

How is practice different for childhood apraxia of speech?

Childhood apraxia of speech needs motor-based practice with a lot of trials per session, often 60 to 100, paired with specific feedback on the movement itself. That's a different approach from typical language-delay practice, which tends to be naturalistic and low-demand. CAS home practice needs to be guided by an SLP who's identified the specific motor targets, because practicing the wrong movements can actually reinforce error patterns. Frequency matters enormously here, which is exactly why parent-implemented practice is so valuable for CAS.

Sources

1. ASHA (American Speech-Language-Hearing Association), Speech and Language Developmental Milestones: ASHA guidance that children who receive parent-implemented home practice between sessions make faster progress than those who only attend weekly sessions
2. Roberts, J. & Kaiser, A. (2011). The Effectiveness of Parent-Implemented Language Interventions. American Journal of Speech-Language Pathology, 20(3), 180 to 199.: Meta-analysis of 18 studies finding parent-implemented language interventions produce statistically significant gains in expressive and receptive language; authors state 'parent-implemented language interventions are effective for improving child language outcomes'
3. ASHA, Speech Sound Disorders: Articulation and Phonology: ASHA developmental norms for speech sound acquisition by age
4. American Academy of Pediatrics, Language Development: Speech Milestones for Babies: AAP criteria for when to refer for speech evaluation including fewer than 50 words or no two-word combinations by 24 months
5. Girolametto, L. & Weitzman, E. (2002). Responsiveness of child care providers in interactions with toddlers and preschoolers. Language, Speech, and Hearing Services in Schools, 33(4), 268 to 281.: Research showing commenting versus questioning produces better language input quality; caregiver responsiveness and modeling quality matter more than raw session minutes
6. Dawson, G. et al. (2010). Randomized, Controlled Trial of an Intervention for Toddlers With Autism. Pediatrics, 125(1), e17, e23.: 2010 RCT in Pediatrics showing ESDM with parent coaching component produced measurably greater language and adaptive behavior gains than community-treatment control
7. Beukelman, D. & Mirenda, P., Augmentative and Alternative Communication (4th ed.), Paul H. Brookes Publishing: Research base supporting dense AAC modeling by communication partners; thirty exposures to a symbol in a single day is not too many
8. American Academy of Pediatrics, Screen Time and Children: AAP recommendation that screen time for children under 18 months be avoided except video chatting, and children 18-24 months only use screens with a caregiver present
9. Pennington, L. et al. (2018). Speech and language therapy for dysarthria in Cochrane Database of Systematic Reviews; broader Cochrane review on early language interventions reporting 8-12 week parent-implemented programs showing stronger effects at higher intensity: Cochrane review evidence that frequency and intensity of parent-implemented practice correlates with language intervention outcomes, with 8-12 weeks of near-daily practice showing stronger effects
10. Apraxia Kids, Treatment for Childhood Apraxia of Speech: Recommendation that children with CAS benefit from high-trial-density practice (60 to 100 trials per session) with high feedback rates based on motor learning research
11. U.S. Department of Education, IDEA Individuals with Disabilities Education Act: Federal mandate under IDEA Part C for free early intervention services for children birth to age 3, including free evaluations; IDEA Part B for school-age children
12. Grogan-Johnson, S. et al. (2011). A Comparison of Service Delivery Models for Speech Therapy with School-Age Children. Communication Disorders Quarterly, 32(4), 200 to 212.: Telehealth SLP services have shown comparable outcomes to in-person therapy for most speech and language goals