Evaluative research has lagged behind the expanding clinical applications of group psychotherapeutic methods. Historical, social and cultural factors, as well as diverse theoretical orientations and a preoccupation with the differences between various group approaches, have contributed to the dilemma of defining group psychotherapy. No generally acceptable definition of what represents improvement in group therapy has yet evolved. A variety of research strategies have been developed to assess group psychotherapy outcome. Approaches using construct criteria, with an individualized outcome for each patient, would seem to offer the most potential for future outcome studies. Many variables may influence evaluative research in group psychotherapy and the conventional research approach of a controlled clinical trial has rather limited feasibility. In discussing the problems of evaluative research, into group psychotherapy outcome, the literature is reviewed.