Last updated 2026-07-10
Pretend play asks a lot of a child's brain all at once: flexible thinking, imitation, and social imagination. For an autistic child, the way in is usually simple object play first, with an adult modeling actions rather than asking for them, following whatever the child already likes, and adding complexity slowly. Structured play intervention has a real, measurable effect on symbolic play in autistic children, and the earlier you start, the more it tends to pay off.
Why this doesn't come easily
Pretend play asks a child to hold two realities at once: this block is a block, and right now it's also a car. That kind of mental flexibility, sometimes called symbolic or representational thinking, is one of the areas autism affects most directly. A 2010 review in the journal Autism Research found that deficits in symbolic play are among the earliest and most consistent markers of autism spectrum disorder, often showing up before age two. [1]
That doesn't mean autistic children can't pretend, just that the path there looks different and needs more scaffolding. Many kids on the spectrum gravitate toward cause-and-effect toys, lining objects up, or repeating scripts from videos. None of that is wrong. It's a starting point.
The second piece is imitation. Typical pretend play grows partly because toddlers watch adults and other kids and copy them without being asked. Research from the MIND Institute at UC Davis has shown that autistic children have reduced spontaneous imitation, meaning they need more deliberate, supported chances to see and practice the actions that eventually become pretend play. [2]
Social motivation matters too. So much early pretend play is social: playing house, playing doctor, running a pretend restaurant. For kids who find social interaction less rewarding or harder to read, the pull toward that kind of play is weaker. So if your child walks away from the tea party you set up, that's not a rejection of you. It just tells you something useful: design play that doesn't depend on social motivation to keep it running.
What the research actually shows
The research base is real but not huge, and most studies use small samples. Still, several well-designed intervention studies point the same direction.
A 2006 randomized controlled trial by Kasari and colleagues in the Journal of Autism and Developmental Disorders found that children who got joint attention and symbolic play training showed significantly more diverse play acts and higher levels of symbolic play than controls. [3] Therapists ran the intervention with children averaging around 36 months, and the effects held at a six-month follow-up.
A 2013 meta-analysis in Research in Autism Spectrum Disorders looked at play interventions across 21 studies. Structured, adult-mediated play instruction produced moderate-to-large effect sizes on symbolic play outcomes. [4] The same analysis found that peer-mediated approaches, where a typically developing peer is coached to play alongside the autistic child, also produced gains, though the evidence there was thinner.
Naturalistic developmental behavioral interventions like JASPER (Joint Attention, Symbolic Play, Engagement, and Regulation), developed at UCLA, have shown strong results across multiple trials. JASPER targets play routines and joint engagement directly. A 2019 trial in JAMA Pediatrics found it significantly improved play diversity and initiations in autistic children. [5]
You don't need to run a clinical trial in your living room, though. The principles behind these interventions transfer straight to daily life: follow the child's lead, keep play joint, build from simple to complex, and use the routines you already have.
Where kids tend to get stuck
Knowing the developmental sequence lets you meet your child where they are instead of dropping something several steps ahead onto the floor.
| Stage | Typical Age | What it looks like |
|---|---|---|
| Exploratory / sensory play | 0-12 months | Mouthing, banging, shaking objects |
| Functional play | 12-18 months | Using objects as intended (cup to mouth, phone to ear) |
| Combinatorial play | 18-24 months | Combining objects (block in truck, stacking) |
| Symbolic play (early) | 24-30 months | Pretending an object is something else (banana as phone) |
| Symbolic play (advanced) | 30-48 months | Multi-step pretend sequences, doll play, role-taking |
| Sociodramatic play | 4+ years | Extended shared narratives with others |
Many autistic children have strong exploratory and functional play but stall at the jump to symbolic play. [1] Some get very good at one specific pretend script, like a memorized scene from a show, but don't generate new pretend sequences of their own. Others manage symbolic play alone just fine and shut down the moment a partner joins. Which gap your child is dealing with changes what you should try next: if your child is still mostly in exploratory play, start there rather than skipping ahead; if functional play is reliable, that's your bridge into early pretend.
How to actually start
Begin with what your child already loves, and not just because it's kinder. Kasari's JASPER research shows that following the child's attentional lead, rather than redirecting them to what you think they should play with, raises engagement and produces better play outcomes. [5] If your kid loves trains, the first pretend play you teach involves trains.
Get into their play without changing it first. Sit at their level, pick up a train, run it on the track next to theirs, and don't talk much. Match what they're doing. This is parallel play, and it's the prerequisite for joint play.
Once they tolerate your presence without leaving, add one action. Make the train go into a tunnel. Make it stop at the station. Make a character fall off. One action, then pause and see if they imitate or respond. Don't demand it.
Next, introduce a character or a voice: a small figure riding the train, a quick "whoa, too fast!" in a character voice. You're adding a layer of symbolism without asking the child to produce it themselves.
Then build a simple routine. Play out the same short sequence several sessions in a row: train loads passengers, train goes, train arrives, passengers get off. Predictability is a feature here, not a bug. It lets the child anticipate, and anticipation is how they start to join in.
After that, leave a gap. Run the routine, then pause before the last step, and wait to see if they fill it. This is the moment pretend play starts becoming theirs instead of yours.
Finally, add substitutions slowly. Put a block in the train car instead of a plastic person. "Look, the block is riding!" Object substitution gets easier over time because you've built a scaffold around it. [3]
None of this comes with a timeline. Some kids move through these steps in weeks. Others take months. Both are fine.
What to hand them
The research crowns no single toy, but the intervention literature gives a few reliable principles.
Realistic props work better early on. A toy telephone teaches the gesture before a banana-as-telephone means anything. Once functional play with realistic objects is solid, you can fade to less realistic substitutes. [3]
Open-ended materials like blocks, cardboard boxes, and fabric are great once the child has some symbolic play under their belt. Before that, they mostly get mouthed or sorted, which is fine, but it won't move pretend play forward on its own.
Water and sand deserve a mention too. Plenty of autistic children who avoid structured pretend play will sink deep into sensory materials, and you can narrate and model pretend scenarios there with almost no pressure.
Dolls and figures split kids hard. Some autistic children find doll play alien and uncomfortable, so don't force it. If your child prefers vehicles or animals, work with those instead. There's nothing magic about a baby doll: what matters is object-plus-action, and a dinosaur does that job just as well.
Some parents find the fastest first step is letting the child watch them play with no requirement to join in. Play nearby, talk to yourself, be interesting. Curiosity often pulls the child in on their own terms.
If your child is a visual learner, a simple picture sequence of a two or three-step play routine posted nearby can help them see what the "script" is. This doesn't cap creativity. It just gives them a predictable structure they can eventually riff on.
How does pretend play connect to language development?
Play and language grow together rather than developing on separate tracks. Both depend on symbolic thinking: a word stands for a thing, just like a block stands for a car. Several studies find that symbolic play level is one of the strongest predictors of expressive language in young autistic children. [1]
The American Speech-Language-Hearing Association (ASHA) notes that speech-language pathologists routinely assess play during a language evaluation, because play complexity and language complexity move together. [6] A child who combines objects in play (cup on the plate, person in the car) is often at the stage where word combinations are starting too, so working on pretend play isn't a detour from language. It's the same road.
For children who use AAC or are minimally verbal, pretend play is still open and still worth teaching. An AAC device can join the play: a symbol for "hungry" becomes what the doll says. A child using aac devices can communicate inside pretend scenarios in ways that build communication and play complexity at once.
If your child uses a lot of scripted language or repeated phrases from videos, that's echolalia, and it can actually be a bridge into pretend play. Many kids first use words from a favorite show to narrate play sequences, and that's a legitimate start. The script becomes theirs over time.
What role do parents play in building this skill?
A big one. Parent-implemented interventions have solid evidence behind them. A 2014 study in Topics in Early Childhood Special Education found that when parents were trained in JASPER strategies, their children's play diversity improved significantly, and the gains matched therapist-delivered sessions. [7]
The parent behaviors the research keeps identifying: get down to floor level, cut back on verbal demands, imitate the child before trying to extend them, and wait. That last one is hard. Most parents fill silence, but in play with an autistic child, a three to five second wait after you model something is often the exact space where they respond. Fill it too fast and you erase the chance.
You don't need structured play sessions. Twenty minutes of intentional play during a natural part of your day, after school or before bath, is plenty. Consistency beats duration.
Go easy on prompting, too. Ask "what does the cow say?" or "where's the baby going?" every thirty seconds and the play turns into a quiz. Your child either shuts down or starts performing for answers instead of playing. Occasional questions are fine. Mostly, narrate and model.
If you're not sure where to begin, a speech-language pathologist who specializes in autism can assess your child's current play level and hand you specific targets. Speech therapy isn't only for articulation: play-based sessions are common and effective at this age, and connecting with early intervention services can often get you that support at no cost if your child is under three.
Can peers help teach pretend play?
Yes, with structure. Peer-mediated play interventions, where a neurotypical peer is coached to stay engaged, offer choices, and follow the autistic child's lead, have shown positive results in several studies. [4] The word that carries all the weight is coached. An adult teaches the peer specific strategies before pairing them with the autistic child. Random peer exposure without coaching tends not to produce play growth, and it can backfire if the peer gives up or takes over.
In school settings, this support can be written into an IEP. If your child is school-age, ask the team about structured peer play opportunities. Social stories and video modeling, where the child watches a video of the target play behavior before trying it, show up often in school contexts with decent evidence behind them. [4]
Siblings can play a similar role, though family dynamics complicate it. A sibling who genuinely enjoys playing with the autistic child is a real asset. A sibling drafted as a therapy tool who resents it is not. Use your judgment.
For autism spectrum speech therapy in a group format, small play groups of two or three children with an SLP facilitating give autistic children a lower-stakes social context than a full classroom.
Does pretend play still matter for older or more advanced kids?
Yes, at least through the preschool and early school years. The symbolic thinking under pretend play is the same thinking that supports narrative understanding, reading comprehension, and social reasoning. Children with richer play repertoires tend to develop stronger theory of mind, the ability to understand that other people hold different thoughts and perspectives. [1]
For older autistic children whose pretend play exists but runs scripted and rigid, the goal shifts from learning to pretend toward expanding play flexibility. Bring in a new character, a new problem to solve in the scenario, or an object that doesn't fit the usual script. That builds cognitive flexibility. It's harder, slower work, and still worth doing.
For autistic teenagers and adults, pretend play as such fades in relevance, but the skills under it don't. Narrative thinking, perspective-taking, and flexible problem-solving stay relevant and can be worked on through improv-style activities, drama, collaborative storytelling, or tabletop role-playing games. Some autistic teens find those games deeply engaging precisely because they come with clear rules and defined roles.
If your child has significant communication support needs alongside limited pretend play, the Little Words app offers a structured, low-pressure space for exploring symbolic language and early play concepts at their own pace, and that daily touchpoint can sit alongside what you're doing with a therapist.
How do I know if my child is progressing, and when do I get help?
Progress in pretend play usually looks small: tolerating a play partner without leaving, imitating one new action, using an object in a new way even once, or staying in a routine longer before disengaging. These aren't dramatic milestones. They're small shifts that stack up.
The American Academy of Pediatrics recommends developmental surveillance at every well-child visit and formal developmental screening at 18 and 24 months using a validated tool. [8] If play concerns are part of what you're seeing, say so plainly. "My child doesn't seem interested in pretend play" is something a pediatrician needs to hear directly.
If your child is under three and not yet doing any functional play (using objects as intended), or if you've been working on pretend play for several months at home with no movement, a formal evaluation is a reasonable next step. Early intervention services for children under three are federally mandated under IDEA Part C and are free to families. [9] The evaluation itself is free whether or not your child ends up qualifying for services.
For children over three, the school district has to evaluate if you request it in writing, and services through the school system are available if the child qualifies. [9]
A speech-language pathologist and a developmental behavioral pediatrician are the two most common starting points. If there's any concern that motor planning is affecting imitation and play, an evaluation for apraxia of speech can be relevant, since apraxia and autism do co-occur and imitation trouble can have a motor component. [10]
Common mistakes parents make
Moving too fast is a big one. Parents catch a spark of interest and immediately escalate: the child picks up a toy phone and the parent launches a full telephone script. The child leaves. Slow down and work one action at a time.
Making it too verbal is another. Play doesn't need narration every five seconds. Too much language overwhelms, especially for children still building language processing. Model silently sometimes. Let the play breathe.
Using toys the child dislikes because they're supposedly "good pretend toys" rarely works. A doctor kit means nothing to a kid who hates it. The child's own interest is the fuel.
Praising every tiny move backfires too. "Great job making the car go! Amazing! You're so good at this!" can interrupt the flow and shift the child's motivation toward earning praise instead of playing. A quiet smile and staying present beats verbal effusion most of the time.
Giving up too soon is probably the most common mistake. Pretend play development in autistic children often takes far longer than parents expect, and that's normal. A realistic window for meaningful change from consistent work is three to six months, not three to six weeks. Nobody has precise data on the average time course for home-taught pretend play, because studies measure group outcomes and individual variation is huge. Keep going.
And don't treat every play moment as a therapy session. Your child needs you to also just be their parent. Some play should carry no agenda at all. That relationship and trust is what makes the intentional sessions work.
Frequently asked questions
At what age should autistic children start pretend play?
In neurotypical development, symbolic play usually shows up around 18 to 24 months. A lot of autistic children get there later, sometimes not until three, four, or beyond. The age itself isn't what matters most. What matters is where your child currently sits in the sequence of play skills and what the next reachable step looks like from there. If something feels off, bring it up at the next well-child visit or reach out to your local early intervention program.
Can autistic children learn pretend play without a therapist?
Yes. Research shows parent-implemented strategies produce real gains in pretend play, and studies on JASPER-trained parents found results comparable to therapist-delivered sessions. Still, a speech-language pathologist or developmental therapist can look at your child's specific play level, give you targeted goals, and notice things you might not. Parents and professionals work best as a team here, not as competing options.
What if my child only does scripted or repetitive pretend play, like re-enacting the same TV scene?
Scripted play still counts as pretend play, and it's a legitimate starting point. Plenty of autistic children get into symbolic thinking through scenes from a favorite show or book. From there, the goal is to stretch it slowly: bring in a new character, change one step, add a small problem that needs solving. Keep the script as your scaffold rather than trying to remove it.
Does teaching pretend play help with language development in autism?
Research keeps showing a close link between symbolic play level and expressive language in autistic children. A 2010 review in Autism Research identified symbolic play as one of the strongest early predictors of later language outcomes. Working on pretend play isn't a detour from language goals, it's often the quickest route to more flexible, functional communication.
How is pretend play different for autistic kids compared to neurotypical kids?
Neurotypical children tend to pick up pretend play on their own, through imitation and social motivation. Autistic children often need the pieces spelled out: what the action looks like, what order it goes in, what the object is standing in for. They may play pretend more when they're alone than with a partner, or more when it's built around their own interest rather than something an adult introduces. The ability is usually there. It just needs a different way in.
What is JASPER and does it work for teaching play?
JASPER stands for Joint Attention, Symbolic Play, Engagement, and Regulation, an intervention developed at UCLA that targets play routines and joint engagement in young autistic children. A 2019 randomized trial in JAMA Pediatrics found it significantly improved play diversity and social communication initiations. Both therapists and trained parents can deliver it, so it's worth asking your child's speech therapist or early intervention team whether they use JASPER or something similar.
My autistic child refuses to play with other children. How do I handle that?
Start with parallel play, sitting nearby and playing without demanding interaction, before expecting joint play to happen. Many autistic children take longer to warm up to a play partner, even a parent. Peer-mediated interventions, where a peer is coached specifically to follow the autistic child's lead, tend to work better than just pairing kids up and hoping for the best. Solo symbolic play should be comfortable first; social play builds on top of that.
Are there specific apps or digital tools that help teach pretend play?
Some apps support narrative and symbolic thinking through story-building, sequencing, or character play, though there isn't much direct research yet on app-based pretend play instruction specifically. Video modeling apps, where a child watches clips of a target play behavior before trying it, have a small but positive evidence base. Any digital tool works best alongside in-person, hands-on play rather than in place of it.
How do I get pretend play support through school or early intervention?
If your child is under three, contact your state's early intervention program: under IDEA Part C, evaluations are free and services cost families nothing if the child qualifies. Once a child turns three, the local school district has to evaluate on written request, and play skills can become part of IEP goals. Ask specifically for a speech-language pathology evaluation that includes play assessment, not just articulation.
Does pretend play matter for autistic kids who are nonverbal or minimally verbal?
Yes. Pretend play is really about symbolic thinking, and that underlies communication no matter how a child communicates. For minimally verbal children using AAC, play gives a natural, motivating reason to use communication symbols: a child who has the AAC symbol for "eat" can use it when the doll is hungry. Play and communication feed each other in both directions.
What's the difference between structured play and free play for autistic children?
Structured play involves adult guidance: you model an action, set up a routine, or present a specific scenario. Free play has no agenda at all. Both matter. Research on pretend play intervention tends to use structured approaches to teach new skills, but free play time, where you just follow the child with no goal in mind, builds trust and motivation and strengthens the relationship that makes structured sessions work in the first place. Try to fit in both over the course of a week.
Can video modeling help an autistic child learn pretend play?
Video modeling, showing a child a video of the target play behavior before they practice it, has a solid track record for teaching social and play skills to autistic children. Several studies, along with a review in Research in Autism Spectrum Disorders, list it among the effective peer-mediated and adult-mediated play strategies. It works best for children who can attend to video, and the behavior being modeled should be simple and close to what the child can already do.
Sources
- Autism Research, Jarrold et al. (2010) – symbolic play in autism review: Deficits in symbolic play are among the earliest and most consistent markers of autism spectrum disorder, and symbolic play level is a strong predictor of language outcomes.
- MIND Institute, UC Davis – autism and imitation research: Children with autism show reduced spontaneous imitation compared to typically developing peers.
- Journal of Autism and Developmental Disorders, Kasari et al. (2006) – joint attention and symbolic play RCT: Children receiving joint attention and symbolic play training showed significantly more diverse play acts and higher symbolic play levels, with effects holding at six-month follow-up.
- Research in Autism Spectrum Disorders, meta-analysis (2013) – play interventions: Structured adult-mediated play instruction produced moderate-to-large effect sizes on symbolic play outcomes across 21 studies; peer-mediated approaches also showed gains.
- JAMA Pediatrics, Kasari et al. (2019) – JASPER randomized trial: JASPER intervention significantly improved play diversity and social communication initiations in young children with autism in a randomized controlled trial.
- American Speech-Language-Hearing Association – autism spectrum disorder practice portal: ASHA notes that speech-language pathologists routinely assess play as part of language evaluation because play complexity and language complexity are closely linked.
- Topics in Early Childhood Special Education, Goods et al. (2013) – parent-implemented JASPER: Parents trained in JASPER strategies produced significant improvements in their children's play diversity, with gains comparable to therapist-delivered sessions.
- American Academy of Pediatrics – developmental surveillance and screening policy: AAP recommends formal developmental screening at 18 and 24 months using a validated tool and developmental surveillance at every well-child visit.
- U.S. Department of Education – IDEA Part C early intervention overview: Under IDEA Part C, evaluations for children under three are free to families and services are provided at no cost if the child qualifies; children over three are entitled to evaluation through the school district upon written request.
- ASHA – childhood apraxia of speech and co-occurring conditions: Childhood apraxia of speech and autism spectrum disorder co-occur, and motor planning difficulties can affect imitation and play.