This study examined 2-to 3-year-olds' ability to make a pretend -real distinction in the absence of content cues. Children watched two actors side by side. One was really eating, and the other was pretending to eat, but in neither case was information about content available. Following the displays, children were asked to retrieve the real food (Experiment 1) or point to the container with the real food (Experiments 2 and 3). 3-and 2.5-year-olds distinguished between the real and pretend acts based on behavioral cues alone. Two-year-olds chose the containers at random, but their spontaneous reactions suggested that they discriminated the real acts from pretense to some degree. Possible accounts for the discrepancy between the different behavioral measures are discussed.Imagine the following situation: A child is observing a person pretending to drink "loudly" from an empty cup. To make sense of this event, the child must interpret the pretend drinking as nonserious, and distinguish it from real drinking behavior. If the child failed to do so, her representations of what real drinking should be like would get confused. Whether, when, and how young children make sense of pretend events is an important scientific issue. Evidence thus far is mixed.Some research suggests that children are able to comprehend others' pretend actions by the age of 28 months, in the sense that they respond correctly to the pretend content (e.g., imaginary tea) and can imagine the outcome of others' make-believe acts (Harris &