This study investigates the effect of curiosity in first and second grade children as an individual difference variable in learning in a computer-based interactive learner control environment and discusses the implications for instructional designers and educators. The instruction was an art education lesson containing both facts and concepts. The study used two learner control treatments: learner control without advisement, in which learners made decisions about content selection, sequencing, pace, remediation, and other issues; and learner control with advisement, in which learners made the same type of decisions but were advised about their options. There was a significant difference in achievement scores in favor of high curious children in both learner control conditions. The results suggest that differences in curiosity influenced performance within either type of learner control lesson. The differences were independent of grade level or gender. There was a significant interaction between grade level and treatment. First grade subjects performed significantly better with advisement, while second grade subjects performed significantly better without. The data indicate that first graders heeded advisement more than second graders. A possible explanation for the high scores of second graders who functioned without advisement may have been the predominance of high curious subjects in that group. (Contains 44 references.) (KRN)