Specific contributions of exploration for divergent and convergent problem-solving ability in preschool children were investigated. Participants were identified as players or nonplayers based on naturalistic observations of free-play activities, and were assigned to either an exploration or nonexploration control condition prior to administration of two divergent and two convergent problem-solving tasks. No significant differences were found between the exploration and nonexploration groups on either divergent or convergent problem-solving tasks. Subjects in the exploration group, however, spent significantly less time in the related convergent problem-solving task than in the nonrelated convergent task. The player/nonplayer distinction was also nonsignificant.