Toddler Pretend Play 2-3 Years: A Research-Backed Guide
Your toddler just spent twenty minutes serving you "soup" made of wooden blocks, narrating "careful, hot" as she set each bowl down. You smiled, played along, and maybe felt a small pang that you should probably be doing something more educational with her. Here's the thing: this is one of the most important things her brain is doing right now.
Pretend play is wildly undervalued by adults. We tend to treat it as filler — the thing kids do when the "real" learning toys are put away. The research tells a different story. Between 24 and 36 months, the imaginary soup, the doll being fed, the cardboard box that is suddenly a rocket ship — these are the visible surface of executive function, language, and social cognition being built underneath.
→ This is a deep-dive into imagination and pretend play. For the full picture across all areas of your toddler's growth — language, fine motor, social-emotional, senses, feeding, and sleep — see our 2-3 Years Main Development Guide.
Why pretend play matters
Key takeaway: Pretend play isn't just fun — research links it to executive function, language growth, theory of mind, and social cognition. The pediatric community now treats unstructured play as a developmental essential, not an extra.
For decades, pretend play was treated as a charming side effect of childhood. The last twenty years of research have changed that view. According to Lillard and colleagues' comprehensive 2013 review in Psychological Bulletin, pretend play is correlated with — and in many studies appears to actively support — several major strands of development at once.
Executive function. When your toddler "cooks" soup, she has to hold a plan in her head, inhibit competing impulses ("the block is not actually food"), and flexibly switch between roles. These are the building blocks of self-regulation — the same machinery she'll later use to wait her turn and focus in preschool. Lillard 2013 reports that children who engage in more pretend play tend to score higher on executive function measures, though the authors note correlation does not equal causation in many of the underlying studies.
Language. Pretend play and language grow together. The narrative structure of "first we cook the soup, then we eat it, then we wash the bowl" maps onto the structure of stories she'll need later. Children with rich pretend play tend to have larger vocabularies and more advanced narrative skills.
Theory of mind. This is the awareness that other people have thoughts and beliefs different from your own. When your toddler hands a stuffed bear an empty cup and says "the bear is thirsty," she is practicing perspective-taking. The Lillard review identifies theory of mind as one of the most robustly linked outcomes of pretend play.
Social cognition. Pretend scenarios require negotiation. Who is the doctor? Whose turn to drive the truck? These tiny conflicts and resolutions are how social skills get rehearsed.
The American Academy of Pediatrics put this all on the official record in 2018. Yogman and colleagues' AAP clinical report The Power of Play makes the case that play — and unstructured pretend play in particular — is essential to healthy development and that pediatricians should be actively prescribing it to families. The report frames play as a buffer against toxic stress and a driver of the executive-function skills that predict later school success.
In other words: the soup is the work. You're not interrupting her learning. You're in it.
The stages of pretend play
Key takeaway: Pretend play emerges in a predictable sequence — functional play around 12-18 months, symbolic substitution around 18-24 months, and sociodramatic play with roles and storylines from about 30 months onward. The ages are ranges, not deadlines.
Pretend play doesn't appear all at once. According to Lillard 2013 and the broader developmental literature, it unfolds in three rough stages — each building on the last.
Stage 1: Functional play (around 12-18 months). Your toddler picks up a toy phone and holds it to her ear. She lifts a toy spoon toward a doll's mouth. She isn't substituting one object for another yet, but she's using objects "as intended" in a pretend frame — the first sign that her brain can hold an idea separate from immediate reality.
Stage 2: Symbolic substitution (around 18-24 months). This is the leap most parents notice. A banana becomes a phone. A block becomes a car. A scarf becomes a baby blanket. Your toddler is using objects to stand in for other objects — a cognitive move that requires holding two ideas in mind at once. Lillard notes that symbolic substitution marks the point when pretend play has truly come online, because it requires the same kind of mental representation that underlies language.
Stage 3: Sociodramatic play (from around 30 months). Pretend stretches into full stories with roles and plots. "You be the doctor, I'll be the baby, and the bear has a sore tummy." It involves negotiation, planning, and improvisation. This is the stage most associated with the executive-function and theory-of-mind gains in the Lillard review.
Between 24 and 36 months, most toddlers move from solid stage 2 into early stage 3. You'll see longer scenarios, more characters, and play that involves you or other children rather than running solo. Don't panic if your toddler is still mostly in stage 2 at 30 months — the range is wide, and sociodramatic play often blooms suddenly between 30 and 36 months.
24-36 month pretend play timeline
Key takeaway: The "normal range" for pretend play milestones is wider than most parents expect — often spanning 6 months or more on either side.
Here's a rough timeline of what pretend play looks like through the third year. Treat these as midpoints, not deadlines. Some toddlers race ahead; others take their time and arrive at the same place by 3.
| Age | Pretend Play Milestone | Normal Range |
|---|---|---|
| 24 months | Uses objects in symbolic substitution (block as phone, banana as car) | 18-28 months |
| 24-27 months | Feeds, dresses, or "puts to bed" dolls and stuffed animals | 20-30 months |
| 27-30 months | Pretend involves a short sequence (cook → serve → eat) | 24-33 months |
| 30 months | Assigns roles to self or toys ("I'm the mommy") | 26-34 months |
| 30-33 months | Pretends with another child in parallel, sometimes together | 27-36 months |
| 33-36 months | Sociodramatic play with multiple roles, a storyline, and negotiation | 30+ months |
| 36 months | Pretend play may involve invisible objects (mimes drinking from empty hand) | 30-40 months |
A toddler at the early end of a range is not behind. A toddler at the late end is not behind. The Lillard review and the broader literature make this point repeatedly — variation is enormous and most "late" pretend players are within the typical range. The thing to watch is forward motion, not whether your toddler hits a specific milestone at a specific month.
The role of props: open-ended beats licensed
Key takeaway: Research consistently finds that open-ended toys (blocks, blank dolls, simple kitchens) produce more language and more elaborate pretend than electronic or licensed toys that "do the work" for the child.
This is one of the most counterintuitive — and best-supported — findings in the toy research. According to Sosa's 2016 study in JAMA Pediatrics, the kind of toy in front of your toddler measurably changes how much you and your toddler talk and how rich that talk is.
Sosa recorded 26 parent-infant pairs playing in 15-minute sessions with three types of toys: electronic toys (a baby laptop, a talking farm, a talking baby cell phone), traditional toys (a wooden puzzle, a shape sorter, rubber blocks with pictures), and books. The result: electronic toys produced fewer adult words, fewer conversational turns, fewer parental responses, and fewer content-specific words than either books or traditional toys. The electronic toys, in essence, did the narrating themselves — and the parents went quiet.
The implication for pretend play is the same. A talking doll that says "I'm hungry, feed me" tells your toddler the story. A blank doll lets her invent the story. A button-laden plastic kitchen that beeps and sings dictates the play. A wooden kitchen with a few pots and pretend foods leaves room for "today the soup is made of leaves."
What this looks like in practice:
- Wooden blocks vs licensed character toys. A pile of blocks can be a house, a car, a wall, a soup ingredient, a baby. A licensed character figure tends to stay that character. Both have a place, but blocks pull more pretend out of your toddler.
- Blank dolls vs talking dolls. A simple cloth or wooden doll lets your toddler assign the doll's mood, hunger, age, and name. A doll that says pre-recorded phrases scripts her.
- Dress-up basket vs themed costume. A bin of scarves, hats, vests, and bags can become anything across a year of play. A single full superhero costume tends to lock in one game.
This doesn't mean you have to throw out every licensed toy in the house. Your toddler will love the character pajamas and the plush dog from the show, and that's fine. The point is to make sure the core of her toy collection is open-ended — the props that can become anything.
Toycycle's toddler collection is heavy on this kind of open-ended pretend gear because that's what holds value through resale: parents buy it, kids play with it for years, and it gets handed down.
Imaginary friends — almost always normal
Key takeaway: Imaginary companions are common, well-studied, and associated with positive — not negative — social and cognitive outcomes. They're not a sign that something is wrong.
If your toddler starts talking about a friend you can't see, the instinct to worry is understandable. The instinct is also, in almost every case, unwarranted.
The most comprehensive work on imaginary companions comes from psychologist Marjorie Taylor. Her 1999 book Imaginary Companions and the Children Who Create Them, drawn from decades of research at the University of Oregon, found that imaginary companions are far more common than parents realize — Taylor's studies estimate that roughly 65% of children up through age 7 have had an imaginary companion at some point. Many parents only learn about the companion years after the fact.
Taylor's research also overturned the older assumption that imaginary friends signal loneliness or social trouble. On average, children with imaginary companions tend to be more — not less — socially competent, with stronger theory of mind and richer narrative skills. They are not confused about reality: ask a 3-year-old whether her imaginary friend can be seen by Mom, and she'll typically tell you no.
When to play along. Almost always. Set a place at the table if she asks. Ask after the friend's day. You're not reinforcing a delusion; you're supporting a sophisticated act of imagination.
When to gently wonder. If the imaginary friend is consistently scary, hurts your toddler, or seems to be replacing real relationships, mention it at the next pediatric visit. These situations are rare. For most families, the imaginary friend is just a delightful, slightly weird passenger for a year or two.
When to support, when to back off
Key takeaway: Your job in pretend play is to scaffold — narrate, follow, add small extensions — not to direct. The research is clear that adult-led play is less developmentally rich than child-led play with a present, responsive adult.
Adults are notorious for taking over pretend play. We mean well: we ask quiz questions, we redirect the story toward something educational, we correct the plot. Research shows this dampens rather than enriches play.
Narrate, don't quiz. When your toddler hands you a "cookie," say "mmm, this cookie is so warm!" — not "what color is the cookie?" Quizzing pulls her out of the pretend frame and into a test. Narration extends the frame and gives her language to absorb.
Follow her lead. If she says the truck is going to the moon, the truck is going to the moon. Don't redirect to "but trucks drive on roads, sweetie." Pretend is allowed to break physics. The cognitive work she's doing is more important than the realism of the scenario.
Add a small extension, then wait. A useful move: take the play she's offering and add one small step. She's feeding the doll? "Oh, I think the baby wants a song after dinner." Then wait — let her take it or leave it. This is the Lillard-style scaffolding the research supports.
Be present, but not the director. The Yogman 2018 AAP report makes clear that the adult role in play is to be a warm, available partner — not a coach. Sit on the floor. Be in the game. But let her run it.
Resist the urge to make it educational. You don't need to slip the alphabet into the doctor's office game. The pretend play is the learning.
When pretend play seems delayed or absent
Key takeaway: If your toddler isn't using objects to pretend by 24 months, mention it at the next pediatric visit. Routine M-CHAT-R/F screening already covers this. Early support helps regardless of what screening turns up.
Most toddlers slide into symbolic pretend play between 18 and 24 months. Some take longer and most arrive on their own timeline. But pretend play is one of the developmental signals pediatricians watch closely, and it's worth knowing what they're looking at.
If you notice your child isn't yet using objects to pretend by 24 months — feeding a doll, pretending a block is a phone, "cooking" at a toy stove — it's worth bringing up at the next pediatric visit. This is one of the things the M-CHAT-R/F, the standard developmental screening tool at the 18- and 24-month well-child visits, asks about. The American Academy of Pediatrics recommends routine M-CHAT-R/F screening at these visits.
What you might notice: no joint attention during play (your child doesn't look at you for a reaction), only repetitive non-pretend play (lining up cars rather than driving them, dumping the same toy out repeatedly), or no symbolic substitution (a banana stays a banana, never becomes a phone).
Trust your gut. Bring it up at the next visit. Early support is helpful regardless of what the screening turns up.
You know your toddler in a way no one else does. If something feels off — about pretend, about play, about anything — your pediatrician would always rather hear from you than not.
Toy categories that fuel pretend play
Key takeaway: The best pretend-play toys at this age are open-ended, durable, and quiet. The categories below are the ones we see hold value, get reused, and produce real pretend.
Here are the six toy categories that, in our experience and in the research, do the heaviest lifting for pretend play between 24 and 36 months.
Play kitchens. Cooking, serving, washing dishes — a play kitchen maps the most familiar part of your toddler's world. A simple wooden kitchen with a few pots and pretend foods can absorb years of play. Skip the ones with batteries and beeping; the silence is the point.
Dollhouses & figures. A dollhouse is a stage for sociodramatic play. Look for figures with neutral expressions and simple clothes — the more open the doll's "personality," the more your toddler can project onto it.
Dress-up. A basket of scarves, hats, vests, capes, and old purses beats a single themed costume by a wide margin. Dress-up sparks identity play — chef, doctor, dragon, grandmother — exactly the stage 3 sociodramatic territory.
Open-ended blocks. Wooden unit blocks, magnet tiles, and stacking sets become anything: a castle, a road, a wall around the dolls. Blocks are the props research keeps returning to as the most language-rich toys in a toddler's room.
Art supplies. Crayons, paper, play dough, and stickers feed pretend play indirectly — by giving your toddler material to make tickets, signs, money, and props for whatever game is unfolding.
Vehicles & ramps. Wooden cars, trucks, trains, and ramps support narrative play — the truck delivers food to the dollhouse, the car crashes and needs the doctor.
You can find safety-inspected versions of all of these in Toycycle's toddler collection. The toys parents most often resell to us are exactly these categories — because they get played with and stay relevant for years.
Frequently Asked Questions
When does pretend play start?
The first signs of pretend play — your toddler holding a phone to her ear, "feeding" a doll a spoon — usually appear between 12 and 18 months. Symbolic substitution (a block becomes a car) comes online around 18-24 months. Full sociodramatic play with roles and storylines typically emerges between 30 and 36 months. The Lillard 2013 review notes that these ages are midpoints, not deadlines, and the range is wide.
Is having an imaginary friend normal?
Yes — overwhelmingly. According to Taylor's 1999 research, roughly 65% of children up through age 7 have an imaginary companion at some point. Children with imaginary friends tend to score higher, not lower, on measures of social skill and theory of mind. Play along. Set a place at the table. You're supporting a sophisticated act of imagination.
Are screens hurting my toddler's imagination?
The AAP's 2016 Media and Young Minds policy recommends no more than one hour per day of high-quality programming for children ages 2-5, ideally watched with a parent. The research concern isn't that screens are evil — it's that screen time displaces the open-ended, hands-on, language-rich play that builds imagination. If your toddler is getting hours of pretend, books, and free play, a moderate amount of co-viewed screen time isn't going to damage her imagination.
What toys help pretend play the most?
Open-ended ones. According to Sosa 2016, traditional toys (blocks, puzzles, simple pretend props) produce significantly more parent-child language and richer play than electronic toys that talk, sing, or light up. The best pretend-play toy collections lean heavily on wooden blocks, blank dolls, play kitchens, dress-up baskets, and vehicles — and lighter on licensed character toys that script the story.
My toddler only plays one pretend scenario over and over — is that normal?
Yes. Toddlers love repetition. Playing "doctor" or "baby goes to sleep" twenty times in a week is a sign she's mastering the script, not a sign of rigidity. Repetition is how the cognitive moves of pretend get rehearsed and consolidated. Over weeks and months, you'll see the same scenario gradually grow more complex — more characters, longer plots, new twists.
How can I encourage pretend play without taking over?
Be present, narrate rather than quiz, and add small extensions rather than redirecting. Sit on the floor, follow her lead, and let the pretend break physics if she wants it to. The Yogman 2018 AAP report frames the adult role as warm partner, not director. Resist the urge to make it educational — the pretend itself is the learning.
Sources
- Lillard AS, Lerner MD, Hopkins EJ, Dore RA, Smith ED, Palmquist CM. "The impact of pretend play on children's development: A review of the evidence." Psychological Bulletin. 2013;139(1):1-34. DOI: 10.1037/a0029321
- Yogman M, Garner A, Hutchinson J, Hirsh-Pasek K, Golinkoff RM; AAP Committee on Psychosocial Aspects of Child and Family Health, Council on Communications and Media. "The Power of Play: A Pediatric Role in Enhancing Development in Young Children." Pediatrics. 2018;142(3):e20182058. pediatrics.aappublications.org
- Sosa AV. "Association of the Type of Toy Used During Play With the Quantity and Quality of Parent-Infant Communication." JAMA Pediatrics. 2016;170(2):132-137. jamanetwork.com
- AAP Council on Communications and Media. "Media and Young Minds." Pediatrics. 2016;138(5):e20162591. publications.aap.org
- Taylor M. Imaginary Companions and the Children Who Create Them. Oxford University Press; 1999.
- American Academy of Pediatrics + Robins et al. — M-CHAT-R/F is the AAP-recommended developmental screening tool at 18- and 24-month well-child visits.