Seventy-two women and 67 men were assigned at random to one of nine weight-loss treatment groups defined by either homogeneity or heterogeneity with respect to gender and degree of overweight at baseline. The hypothesis tested was that individuals in groups whose members were similar on one or more characteristics would be more successful in weight loss than those with dissimilar group members. Results did not confirm the experimental hypothesis. Although men lost more absolute weight than women and heavier individuals more than lighter, treatment group composition was not related to treatment success. Moreover, analyses controlling for initial body weight showed no gender or initial weight-status differences. The only variable predicting outcomes in the latter analysis was whether or not participants had previously been in a weight-loss program. Those with prior unsuccessful experiences were less successful than those for whom the current study was their first formal weight-loss attempt.