Lead
"I think our child has an invisible friend..." More than a few parents say this with a slightly puzzled look. The companion has a name. The child insists on setting a place at the table. Sometimes the invisible friend is invoked as an explanation for behavior — "Momo is angry, that's why." Faced with all this, parents tend to worry in one of two directions: either something is wrong, or the child must be unusually gifted.
Neither reading fits. Imaginary companions are a common feature of early childhood, and research associates them with positive indicators of social and cognitive development. At the same time, those associations are modest — having an imaginary companion does not make a child exceptional. This article sets out what the research actually says, without inflation in either direction.
How common are they?
Prevalence figures vary considerably depending on how "imaginary companion" is defined.
Using a narrow definition — an invisible person or creature the child treats as a friend — roughly 20 to 30 percent of preschool-age children qualify. Expand the definition to include personified objects (stuffed animals or toys invested with a distinct personality and relationship) and the figure rises to somewhere between 40 and 65 percent [1].
Taylor and Carlson (1997) studied 152 children ages three and four. Using the broader definition, approximately 63 percent had either an imaginary companion or a special relationship with a personified object [2]. Taylor's 1999 book, the most comprehensive treatment of the phenomenon, cites roughly 65 percent — an estimate that combines invisible companions and personified objects [1].
It is also worth noting that this is not an on/off phenomenon. Imaginary companionship sits on a spectrum of pretend play and fantasy activity. Children who seem to "not have one" often engage in related forms of imaginative play that share the same underlying cognitive features — talking to toys as if they have feelings, enacting extended narratives with small figures, or maintaining an elaborate story world that persists across play sessions. The boundary between these activities and a "true" imaginary companion is not crisp, and the 63 to 65 percent figure should be read as a rough indication of how widespread the imaginative territory is, not as a precise census.
The connection to theory of mind
The most consistent finding in this literature is that children with imaginary companions tend to perform better on theory of mind: the cognitive capacity to understand that others have beliefs, desires, and intentions distinct from one's own tasks than peers without them.
Taylor and Carlson (1997) found that among four-year-olds, fantasy play (including imaginary companionship) was significantly associated with success on false-belief tasks and appearance-reality distinction tasks — even after controlling for verbal ability [2].
A three-year follow-up by Taylor and colleagues (2004) showed that the association persisted into early elementary school: children who had engaged in more fantasy play as preschoolers showed stronger scores on measures of emotional understanding and social cognition [3].
The important caveat, as with any correlational finding, is that the direction of causation is unknown. Does having an imaginary companion build theory of mind? Does a more developed theory of mind make rich imaginary companionship possible? Or does some third factor — perhaps a general capacity for symbolic representation — support both? The research cannot answer this. Leslie's (1987) metarepresentation: the ability to hold a mental state that is decoupled from reality — representing another representation rather than directly encoding the world theory offers the most internally consistent account: both pretend play and theory of mind may draw on the same underlying ability to represent mental states [4]. But that remains a theoretical framework, not a settled fact.
Gender, sibling status, and the limits of those associations
Imaginary companions are reported more frequently among girls and among only children or children with few siblings [5]. Hoff (2005) found both patterns. The sibling finding carries a plausible developmental interpretation: a child who spends more time without ready play partners has more occasion to create social interaction through imagination, and a companion that the child controls fully may serve as a kind of practice space for social exchange.
These associations, however, are inconsistent across studies, and the effect sizes are small. Hoff's own work situates imaginary companionship within a broader picture of creativity and self-image in middle childhood — neither a deficit signal nor a straightforwardly positive marker, but part of the wider variation in how children use imaginative resources [5]. The sibling finding in particular risks being read through an inaccurate narrative — that only children are lonely, and the imaginary companion compensates for isolation. The data do not support that framing. Imaginary companionship is not a symptom of social deprivation; it appears across family configurations and alongside typically developed peer relationships.
How to respond
Research consistently points to one thing parents should not do: dismiss or correct the imaginary companion. Responses like "there's no such person" or "stop making things up" tend to close down a child's imaginative space and narrow the communication between parent and child.
That said, parents are not required to treat the imaginary companion as literally real. When a child says "Momo needs dinner too," a parent does not need to agonize over whether to set another place. Going along lightly — "oh, Momo's joining us?" — while letting the child lead the imaginative world is the natural middle ground.
Most imaginary companions fade around age seven, roughly when robust real-world friendships through school become available. There is no need to worry about the departure, and no reason to hurry it along either.
A note on recording: logging exchanges involving an imaginary companion — in an app like Memori or in any diary — preserves a particular kind of childhood document. Years later, these fragments of a child's inner world at three or four are the hardest to reconstruct from memory alone, and unlike photographs, they capture the texture of language and imagination at a specific moment.
Is this a symptom of a mental health problem?
Some parents worry that an imaginary companion is a sign of psychological difficulty. The research evidence runs in the opposite direction: in typically developing children, imaginary companions are associated with positive mental health indicators, not negative ones [1,2].
This is distinct from a different phenomenon: a child reporting that voices are giving commands, or that an entity is controlling behavior in a distressing way. These warrant professional consultation. The distinguishing question is whether the child seems to enjoy the relationship and whether it is interfering with daily life. A child who cheerfully describes Momo's preferences and can pivot back to ordinary reality without distress is not showing a clinical sign. If a parent is genuinely uncertain, a conversation with a pediatrician is always appropriate — but uncertainty alone is not cause for alarm.
Summary
Imaginary companions appear in a substantial proportion of preschool-age children — between 20 and 65 percent, depending on how broadly the phenomenon is defined [1]. They correlate positively with theory of mind performance, but the causal direction of that relationship is unresolved, and the correlation does not justify characterizing children with imaginary companions as developmentally advanced [2,3].
The practical response is to neither dismiss nor over-literalize the companion. Let the child lead the imaginative space. The imaginary companion will most likely fade on its own as school-age friendships develop. It is not a warning sign. It is not a certificate of talent. It is the imaginative capacity of an early childhood working at full stretch.
References
- Taylor M. Imaginary Companions and the Children Who Create Them. New York: Oxford University Press; 1999.
- Taylor M, Carlson SM. The relation between individual differences in fantasy and theory of mind. Child Dev. 1997;68(3):436–455. doi:10.1111/j.1467-8624.1997.tb01950.x. PMID: 9249959.
- Taylor M, Carlson SM, Maring BL, Gerow L, Charley CM. The characteristics and correlates of fantasy in school-age children: imaginary companions, impersonation, and social understanding. Dev Psychol. 2004;40(6):1173–1187. doi:10.1037/0012-1649.40.6.1173. PMID: 15535765.
- Leslie AM. Pretense and representation: the origins of "theory of mind." Psychol Rev. 1987;94(4):412–426. doi:10.1037/0033-295X.94.4.412.
- Hoff EV. Imaginary companions, creativity, and self-image in middle childhood. Creat Res J. 2005;17(2–3):167–180. doi:10.1080/10400419.2005.9651477.