Although the literature abounds with speculation about open education, there are few experimental studies designed to test its effects. This research compared an open education program with a traditional program at the middle school level. Subjects were randomly assigned to comparison groups. Dependent measures included a self-concept scale, a measure of creativity, locus of control, and five achievement measures. Pre-test intellective measures showed equivalence between groups and post-test analysis found only one difference (creativity) which favored open education. Implications for "fitting" open education goals to curriculum programming are discussed. (Author)