The hypothesis (Kelly, 1968) that, in a fluid school setting, students with a high preference for exploration would adapt more easily than low explorers, while in a constant school environment the opposite would be the case, was examined. A sample of 120 eighth-grade boys selected to represent low, moderate, and high exploration groups were followed as they progressed through two large suburban high schools. The first high school (with a population exchange rate of 19%) had a relatively fluid social environment, while the second high school (with a 7% rate of population turnover) had a relatively constant social environment. Two-way analysis of variance found significant School x Exploration Interaction in the hypothesized direction for twelfth-grade measures of psychosomatic symptoms of self-dissatisfaction. An investigation of assumptions found that in the more fluid setting high explorers reported a greater degree of affective interaction and instrumental interaction with peers and school staff than did low explorers, but in the more constant setting this was not the case. The implications of these findings for school intervention strategies were discussed in terms of a theory of person-environment interaction. (Author)