The purpose of this paper was fourfold. The first purpose was to demonstrate the need to develop an instrument to assess group cohesion while the second was to outline a conceptual model of group cohesion upon which such an instrument could be based. This model reflected four related constructs which were the a priori basis for developing a large item pool and initial versions of the Group Environment Questionnaire (GEQ). The third purpose was to outline the four projects conducted to obtain construct-related information and to develop an initial version of the GEQ. The final purpose was to outline the two reliability and validity studies conducted with two different sport team samples. The results of these studies revealed that an 18-item version of the GEQ was internally consistent, reliable across studies, and content valid. Factor analyses with oblique rotation revealed preliminary evidence for construct validity. The GEQ is comprised of four scales reflecting the constructs of group integration-task, group integration-social, individual attractions to group-task, and individual attractions to group-social.