Twenty-four girls and 24 boys enrolled in a university preschool were given two trials each to perform a marble-dropping task. After the first trial, subjects in the sex-appropriate condition were told that children of their sex perform the task better than children of the opposite sex. Subjects in the sex-inappropriate condition were told that children of their sex perform the task less well than children of the opposite sex. It was hypothesized that sex-appropriate subjects would set higher goals for Trial 2 than control subjects receiving no information and that sex-inappropriate subjects would set lower goals than control subjects. The hypotheses were not supported. Instead, children of both sexes set significantly higher goals when told that boys do better at the task. Interpretations of the results are offered.