This design allowed us to simultaneously address our three major questions. First, the comparison between information in pretend play and non-generic language allows us to assess whether pretend play offers an advantage in supporting children’s learning of general knowledge. Second, because the information always concerned familiar animals rather than novel ones, finding that children learned general knowledge will show that pretend play supports learning about familiar kinds. Finally, because the test question does not prompt children to think about kinds, if children give generic responses, this will show that they generalize information from pretend play without such prompting.