The purpose of this study was to compare aspects of educational environments in elementary schools which differed in the degree to which programs were individualized flexible, and diverse; and architectures were open. Fifteen classes were assigned to the cells of a factorial design, and data on educational environments were collected from teachers and students by systematic observations, interview, and self-report procedures. Multivariate analyses of variance were conducted using both individual and aggregated scores. The study demonstrated that program openness is useful in distinguishing among educational environments. The empirical clarification of the concept was the major result of the study. (Author/MLF)