This study examined how rumors originating in 3- to 6-year-olds' causal inferences can affect their own and their peers' memories for a personally experienced event. This was accomplished by exposing some members of classrooms to contextual clues that were designed to induce inferences about the causes of two unresolved components of the event. After a 1-week delay, a substantial number of children who were exposed to the clues misremembered their inferences as actual experiences. Causal inferential memory errors were most pronounced among 5- and 6-year-olds. Also, many of the children whose classmates were exposed to the clues mistakenly incorporated their classmates' causal inferences into their own accounts, with 3- and 4-year-olds being most likely to make this error.