Last updated 2026-07-11
TL;DR
Incidental teaching means waiting for your child to want something, then asking for a little more communication before you hand it over. Studies show it increases spontaneous language in children with autism and other language delays, and it works especially well when parents are the ones doing it. You can use it at meals, during play, and on errands. No clinic required.
What incidental teaching actually is
Incidental teaching is a structured but natural way to build communication, developed in the early 1970s by researchers at the University of Kansas [1]. Rather than sitting your child down for drills, you wait until they want something, then use that moment of real motivation to ask for a bit more language before handing the thing over.
It's child-led in a specific sense: you don't pick the topic, your child does, by reaching, pointing, looking, or making a sound toward something they actually want. Your job is to set up the room so those moments happen often, then respond in a way that asks for slightly more than what your child just gave you.
The technique belongs to a larger family called naturalistic developmental behavioral interventions (NDBIs). The American Speech-Language-Hearing Association (ASHA) treats NDBIs as an evidence-based practice for children with autism spectrum disorder and other language disorders [2], which sets it apart from the endless supply of therapy hacks online that rest on nothing.
A word on terms: some sources use "incidental teaching" and "milieu teaching" almost interchangeably, but they're close cousins rather than twins. Milieu teaching is the umbrella term, and incidental teaching is one procedure under it, the one focused on moments the child starts rather than ones the adult arranges [3].
Does it actually work?
The evidence is good but not huge. Most studies are small, which is normal for behavioral speech research. Still, the results line up well enough that ASHA lists incidental teaching as a recommended practice.
A meta-analysis by Goldstein in the Journal of Early Intervention reviewed naturalistic language studies and found that milieu teaching procedures, including incidental teaching, produced reliable gains in spontaneous word use and carried those words into new settings [4]. That carryover is the whole point: kids don't just say a word in the therapy room, they use it at home, at the park, with grandparents. A 2014 study in the Journal of Autism and Developmental Disorders looked at children with ASD and found that naturalistic teaching produced stronger generalization of new words than discrete trial training on its own [5]. Neither approach is magic; kids who got a mix of structured and naturalistic teaching generally did best.
For late talkers without an autism diagnosis, the evidence is thinner, and nobody has strong randomized trial data here yet. The closest support comes from parent-run naturalistic interventions, several of which show real gains in early vocabulary when parents are trained to use the techniques consistently at home [3].
The four steps
Four steps, in order, often finished in under 30 seconds. Once the rhythm clicks, it stops feeling like a script.
First, arrange the environment. Put interesting things in view but out of reach: bubbles on a high shelf, a favorite snack in a clear container they can't open, a toy with pieces missing. You're building opportunities, not building the child's interest for them, since that part is already there.
Second, wait and watch, which is the hard part. Do nothing. Make eye contact, stay present, and wait for your child to signal they want something, whether that's a reach, a point, a sound, a word, or a long look at the object. You're waiting for the child to start it, so don't jump in early.
Third, prompt for more. Once they signal, ask for communication one level above what they gave you. If they reached silently, prompt for a sound or a point. If they pointed, prompt for a word approximation. If they said "ba," prompt for "ball." If they said "ball," prompt for "red ball" or "want ball." One step up, never five; this is sometimes called a least-to-most prompting hierarchy. Common prompts run from lightest to heaviest: an expectant look (time delay), a partial model ("ba..."), a full model ("say 'ball'"), or a question ("What do you want?"). Start light, and if they respond, you're done. If they don't within 3 to 5 seconds, move to a more direct prompt. Fourth, respond and deliver. The second they give you any communicative attempt, even a rough one, name it and hand it over: "Ball! Here's your ball." The reward is natural (they get the thing they wanted) and it's instant. Don't withhold the item to squeeze out more. One prompt cycle, then deliver.
How it compares to discrete trial training
Discrete trial training (DTT) is the best-known ABA technique: you present a stimulus, the child responds, you give feedback, and you repeat. It's tightly structured, adult-directed, and usually happens at a table. Incidental teaching flips that around. The child starts it, it happens in natural settings, and the reward is the actual thing they wanted, not a token or a piece of candy. The two aren't rivals, and plenty of speech therapists and behavior analysts use both.
| Feature | Incidental Teaching | Discrete Trial Training |
|---|---|---|
| Who initiates | Child | Adult |
| Setting | Natural environment | Structured table work |
| Reinforcer | The requested item/activity | External (token, praise, edible) |
| Generalization | Strong | Requires extra programming |
| Best for | Spontaneous language, generalization | Teaching new concepts, building foundational skills |
| Session feel | Conversational | Drill-like |
If your child is at the earliest stages of communication, DTT might build the first skills that incidental teaching then stretches into real life. Your child's speech therapist can tell you which approach fits where your child is right now.
Working it into the day
Basically all the time, but in short bursts. This isn't a scheduled 20-minute session, it's a lens you put on ordinary moments. Mealtimes are gold, since food is highly motivating: snacks in a closed container, a drink left just out of reach, a favorite food cut into small pieces so your child has to ask for more. Each one is a built-in opportunity. Play works too. Bubbles where you hold the wand, blocks where you hold the pieces, a wind-up toy your child can't crank alone. You become the assistant they have to talk to. Getting dressed, bath time, car rides, grocery runs: each has predictable words (shoes, shirt, water, more, help, out) and natural moments of wanting things. Repeating those words across settings is what builds carryover. One limit worth keeping in mind: if your child is dysregulated, tired, sick, or clearly finished, stop. Incidental teaching needs a child who is motivated and calm enough to try. Pushing through a meltdown teaches nothing, and it can make communication feel like a chore worth avoiding.
Setting up the environment
This is called environmental arrangement, and it's half the strategy. You can't wait for a child to start something if nothing interesting is around to start about. Out-of-reach items work well: preferred toys on a visible shelf, snacks in see-through containers at eye level but out of reach. If your child can grab whatever they want, there's no reason to communicate. Sabotage routines help too. Give them a cup with no juice in it. Zip a jacket most of the way and stop. Put shoes on the wrong feet, but only if your child thinks it's funny rather than upsetting. These small "mistakes" open the door to communication. Incomplete sets do the same job: playdough with no tools, a puzzle with pieces held back, crayons and no paper. When the expected piece is missing, many kids will signal for it. Sometimes the setup is just you. Sit across from your child during play, hold an item they want, and wait with a warm, curious face. Your expectant look is itself a cue that says communication is expected. Skip cluttered spaces during focused practice. Forty toys on the floor means your child never needs you. Fewer choices, more communication.
What communication levels can incidental teaching target?
Parents sometimes assume this approach only works once a child already has words. That's not true. It scales to wherever your child is right now.
For pre-verbal children or those using gestures only, you target any intentional communication: a reach, a point, eye contact paired with a sound, handing you an object. You're not waiting for words, you're teaching the idea that communication makes things happen.
Once a child has single words, the target shifts to two-word combinations like "more juice," "want ball," or "daddy help." At two to three words, you push toward longer phrases, pronouns, prepositions, and questions.
Kids using AAC devices or picture systems follow the same basic pattern: wait for interest, prompt them to use their device or board to request, then deliver what they asked for. The American Academy of Pediatrics supports early AAC use and notes it does not slow speech development [6]. If your child uses AAC devices, this kind of teaching is one of the best ways to move them from copied, prompted-only use toward real, spontaneous requests.
If your child has echolalia, adapt with a lighter touch. When they echo your model back as a working request, that counts as a real communicative attempt and deserves a reward. Don't hold out just because it wasn't spontaneous.
How do you prompt without turning it into a fight?
This is the balancing act at the center of incidental teaching: requiring communication without making every snack a standoff that ends in tears.
Keep the wait short. Three to five seconds of expectant silence is usually plenty before you prompt; drag it out to 30 seconds and most kids escalate or check out entirely. Meet your child where they actually are: if they're working hard and give you "buh" for "bus," take it. Rewarding effort beats holding out for perfect articulation, especially early on. Never end on a failure. If they don't respond to your prompt, drop back to a full model, let them imitate or not, and hand over the item anyway with natural language alongside it. The interaction should end on a good note, not a battle. Watch for signs you've aimed too high, too: steady refusal, avoidance, or upset the moment you prompt usually means it's time to drop the target down a notch.
And stay warm outside the teaching moments themselves. When your child communicates anything, respond naturally and with genuine interest, even when it's not a "teaching" opportunity. That warmth is what makes them want to keep communicating. A 2001 study in the Journal of Speech, Language, and Hearing Research found parental responsiveness was one of the strongest predictors of language growth in late talkers, regardless of which intervention was used [7].
Do you need a therapist, or can parents do this at home?
Parents can absolutely run this at home, and in a lot of cases parent-led practice beats clinic-only sessions, simply because of the math. A speech therapist might see your child for 30 to 60 minutes once or twice a week. You have hours every day with them. The number of practice moments at home dwarfs anything a clinic can offer.
A 2013 study in the American Journal of Speech-Language Pathology on parent-implemented enhanced milieu teaching found that training parents in naturalistic communication strategies produced significant gains in child communication, with effects comparable in size to therapist-delivered intervention for many children [3].
That said, training still matters. Watching a couple of videos and winging it is harder than it looks: the timing of prompts, how hard to push, when to back off, all of that takes practice. Most speech-language pathologists who work with young kids will coach you during sessions if you ask them to. Just say it directly: "Can you show me how to do this at home between sessions?"
If you're waiting on early intervention services or haven't found a therapist yet, tools like the Little Words app can help you build the habit of language-rich, prompt-and-respond routines in the meantime. Think of it as a habit builder, not a stand-in for a professional assessment. And for families navigating autism spectrum speech therapy, the research points to parent coaching in naturalistic strategies like this one as a core piece of early intervention [9].
What if it isn't working?
Start by checking the prompt level. If your child is pre-verbal and you're holding out for two-word phrases, nothing is going to happen. Know where they actually are and aim exactly one step above it.
Then ask whether the reward is really motivating. Food usually works, but some kids care more about social play, movement, or sensory input. If they don't want the item, they won't work for it. Also look at the environment: if your child can get the thing another way (someone else hands it over, or they can just reach it), the whole setup falls apart.
Some children have motor speech disorders that make words hard to produce on demand. For kids with apraxia of speech, prompt pressure can actually raise the error rate, and an SLP familiar with childhood apraxia of speech may need to change the approach substantially or pair it with other motor speech methods.
Last, check your own execution. Are you waiting long enough before prompting? Stacking prompts too close together? Getting the reward to them quickly after the attempt? Filming yourself during a practice session and watching it back is genuinely useful here. Most people find they're jumping in too fast.
How long before you see results?
There's no clean controlled timeline for this, and anyone promising a set number of weeks is overselling it. What the research does show is that consistency matters more than anything else.
In studies of parent-run naturalistic interventions, researchers usually see measurable language gains within 12 to 16 weeks of consistent use, meaning parents practicing multiple times a day [3]. That's the part that matters: multiple times a day. One or two chances at breakfast won't move the needle. Ten to 20 embedded chances spread across a full day will.
For some kids, especially those whose understanding runs ahead of their talking, you'll notice engagement and comprehension shift before new words actually show up. That's a real sign of progress, not a failure.
Early gains tend to show up in spontaneous requesting, which is exactly what this approach targets. Broader growth, commenting, asking questions, social language, usually takes longer and often needs other strategies layered on top.
If you've been consistent for three months with nothing to show for it, that's worth a conversation with a speech-language pathologist. It might mean the approach needs adjusting, the target level needs rethinking, or something else (hearing, motor speech, sensory processing) deserves a closer look.
Parents often ask when incidental teaching is appropriate to start, and the honest answer is: earlier than you'd think, and later than you'd think. Researchers have used it with children as young as 18 to 24 months, and it's most common during the preschool years, but there's no upper age limit. Older children, teens, and even adults with developmental disabilities have been supported using these same milieu teaching principles. What matters isn't the child's age. It's their current communication level and what actually motivates them. Incidental teaching did grow out of applied behavior analysis, and it uses behavioral tools like reinforcement and prompting hierarchies, so in that sense it's a naturalistic ABA procedure. But because it happens during everyday moments led by parents or therapists rather than at a table, it feels nothing like traditional ABA sessions. Many clinicians fold it into broader ABA or speech therapy programs rather than running it as a standalone approach. It works for non-verbal children too. You're not waiting for words here: you're watching for any intentional signal, a reach, a point, eye contact, a sound, handing you an object, or pressing an AAC button, and then prompting the next step up. Rewarding every one of those attempts teaches a child that communication gets results, long before words show up. Most of the research behind this points to 10 to 20 opportunities a day, spread across ordinary routines. That number sounds bigger than it feels once you realize it's spread across meals, dressing, play, and errands. Quality matters more than hitting a count: each trial needs genuine interest from your child, a clear prompt, and an immediate reward for any attempt. Trials that feel forced or joyless just don't produce the same results. It's worth knowing how this differs from responsive interaction, which is simply following your child's lead and expanding on whatever they offer, without asking for more. Incidental teaching adds that extra step: you wait for your child to initiate, then ask for a slightly more complex response before handing over what they want. Both approaches are child-led, but incidental teaching pushes a bit harder for the attempt. If your child uses PECS, incidental teaching fits right in, since the later phases of PECS already borrow many of the same techniques. Once picture exchange is reliable, you can use incidental teaching moments to build toward longer sentence strips or more spontaneous picture use instead of relying on prompts every time. Echolalia is trickier but not off-limits. If your child echoes your model as a genuine request (you say "juice," they echo "juice" while reaching for the cup), treat that as real communication and reward it. Spontaneous requests tend to increase over time. For children with heavy echolalia, though, loop in a speech-language pathologist first to understand what the echolalia is actually doing before you layer any strategy on top. You don't need to be a trained therapist to do this at home, but coaching helps enormously. Parents who get direct feedback from an SLP, watching them demonstrate it and then trying it themselves, tend to get noticeably better results than parents working from articles or videos alone. Even two or three coaching sessions can sharpen how accurately you use the technique. If your child gets upset when you don't hand something over right away, keep the ask small and back off the moment distress builds. A brief expectant pause of 3 to 5 seconds is worlds apart from holding out for 30 seconds. Frequent meltdowns usually mean the target's too hard or the motivation too weak, so give the item with a natural language model attached and try again later when things are calmer. This should feel warm, never like a standoff. Figuring out the right target takes some judgment. An SLP can assess your child's expressive and receptive levels and tell you exactly what to aim for next. Without one yet, a rough guide works: if your child has no intentional communication yet, reward any signal; if they gesture, ask for sounds; if they vocalize, ask for words; if they use single words, ask for two-word phrases. For most children with real language delays, incidental teaching won't replace formal speech therapy, and it isn't meant to. A proper evaluation figures out what's driving the delay (motor, processing, hearing, social communication), and incidental teaching becomes the high-repetition way you practice whatever targets the therapist sets. Clinic sessions plus home practice consistently beat either one alone. As for where to actually do this: snacks and meals are the easiest starting point since food motivation runs high and the vocabulary is predictable. Bath time (water, soap, in, out, more) and getting dressed (shoes, shirt, on, off, help) work well too, and toys that need a grown-up's help, like bubbles, balloons, or wind-up toys, create constant natural openings. Grocery trips and park visits round things out and help your child use these skills in new places, not just at home.Sources
- Hart & Risley (1975), Journal of Applied Behavior Analysis, original incidental teaching paper: Incidental teaching was developed by researchers Betty Hart and Todd Risley at the University of Kansas in the early 1970s as a naturalistic language intervention
- ASHA, Evidence-Based Practice research portal: ASHA recognizes naturalistic developmental behavioral interventions (NDBIs), including incidental teaching, as evidence-based practices for children with ASD and language disorders
- Kaiser & Roberts (2013), American Journal of Speech-Language Pathology, parent-implemented enhanced milieu teaching: Parent-implemented naturalistic language interventions produced significant improvements in child communication outcomes, with effects comparable to therapist-delivered intervention; gains seen within 12 to 16 weeks of consistent implementation
- Goldstein (2002), Journal of Early Intervention, meta-analysis of communication intervention for children with autism: Milieu teaching procedures including incidental teaching produced reliable gains in spontaneous word use and generalization of language to new settings
- Kasari et al. (2014), Journal of Autism and Developmental Disorders, naturalistic vs. structured intervention: Naturalistic interventions for children with ASD produced stronger generalization of new words compared to discrete trial training alone; combined approaches generally produced the best outcomes
- American Academy of Pediatrics, healthychildren.org: The AAP supports early AAC use and notes it does not impede speech development in children with autism or language delays
- Yoder & Warren (2001), Journal of Speech, Language, and Hearing Research, parental responsiveness and language growth: Parental responsiveness was one of the strongest predictors of language growth in late talkers, independent of intervention type
- ASHA, Augmentative and Alternative Communication (AAC) practice portal: AAC use supports communication development; naturalistic implementation strategies like incidental teaching are recommended for building spontaneous functional AAC use
- National Institute on Deafness and Other Communication Disorders (NIDCD), autism spectrum disorder and communication: Early naturalistic communication interventions are recommended for children with ASD; parent training is a key component of effective early intervention
- Hancock & Kaiser (2006), Topics in Early Childhood Special Education, enhanced milieu teaching overview: Enhanced milieu teaching, which includes incidental teaching as a core procedure, is described as suitable for children across communication levels from pre-verbal to early phrase users
- Centers for Disease Control and Prevention (CDC), Learn the Signs. Act Early.: Early intervention services including speech-language therapy are federally supported for children birth to age 3 with developmental delays under IDEA Part C