This paper examines the development of cohesion and self-disclosure during the mandatory group experience of a masters level training program in counseling. Differences between cohesion and self-disclosure in these groups, as opposed to what might be expected in a more naturalistic setting, are described and implications for the training of group therapists are discussed. This is a descriptive study based on self-report data of individual members, 21 counseling students in a required group psychotherapy class. They were assigned to four small groups in a quasi-random fashion, attempting to have mixed-gender groups; neither age, ethnicity, nor other factors were considered in distributing group members. Groups met weekly for nine weeks. Every two weeks participants received a combined cohesion and self-disclosure questionnaire consisting of 17 items comprising a cohesion scale, a self-disclosure scale, four distractor questions, and an open-ended question. Results of this study suggest that it is probably wise to insist that leader-trainees subsequently also experience a more typical sample of in vivo group development while under supervision. A copy of the self-disclosure questionnaire is included. (Contains 18 references.) (MAH)