Even though behavioral weight loss interventions are conducted in groups, a social contingency (SC) paradigm that capitalizes on the social reinforcement potential of the weight loss group has never been tested. We tested a weight loss intervention in which participation in the weight loss group was contingent upon meeting periodic weight goals. We hypothesized that making access to the group dependent upon weight loss would improve weight outcomes. Participants (N = 62; 84% female; 94% white; age = 51.9 ± 9.0; BMI = 34.7 ± 4.5) were randomized to 6-months of standard behavioral weight loss (SBWL) or to a behavioral program that included a SC paradigm. Both groups engaged in social cohesion activities. Participants in SC who did not meet weight goals did not attend group meetings; instead, they received individual treatment with a new interventionist and returned to group once their weight goals were met. SC did not improve overall weight loss outcomes (SC: -10.0 ± 4.9 kg, SBWL: -10.8 ± 6.4 kg, P = 0.63). Similarly, overall weight loss was not significantly different in the subgroup of participants in the SC and SBWL conditions who did not meet periodic weight loss goals (-7.3 ± 4.1 kg vs. -7.1 ± 3.5 kg, P = 0.90). Surprisingly, "successful" SC participants (who met their weight goals) actually lost less weight than "successful" SBWL participants (-12.4 ± 3.2 kg vs. -14.5 ± 4.7 kg, P = 0.02). Whereas contingency-based treatments have been tested for other health behaviors (e.g., substance abuse), this is the first study to test a SC intervention for weight loss. This approach did not improve overall weight loss outcomes. Our attempt to offer appropriate clinical care by providing individual treatment to SC participants when needed may have mitigated the effects of the SC paradigm.