This study assesses the effects of visual and verbal rehearsal on the problem solving efficiency of preschool children. The study examines the use of stimulus pretraining designed to induce verbal rehearsal and compares the effectiveness of induced verbal rehearsal with the effectiveness of induced visual rehearsal. Fifty-four middle class preschool children with a mean age of 61.4 months were assigned to one of three conditions: (1) verbal rehearsal, (2) visual rehearsal, and (3) control. Children in each condition participated in a stimulus pretraining session followed by a concept discovery task presented in the blank-trials format. Results indicated that visual rehearsal led to greater solution performance and more hypothesis testing than did verbal rehearsal, and to the use of hypothesis-systems rather than position-dominated response patterns across blank-trial probes. It is suggested that the visual rehearsal procedure used in the training sessions provides a useful operational definition of the visual rehearsal process. (Author/BD)