Small group social skills training (SST) is recommended to improve the social behavior and peer relations of rejected children, but child responses vary. This study explored variation in intervention experiences that emerged in a large study of a school-based SST program for peer-rejected children and associations with program outcomes. One hundred and seven peer-rejected elementary students (60% White, 15% Black, 22% Latinx, 3% Multi-racial, and 65% male) were identified by classroom sociometrics and randomly assigned to a year-long small group SST program. Normative classmates served as partners. Among the peer-rejected targets of the intervention, those with relatively better pretreatment social skills received more SST sessions and were more positively engaged, whereas those with fewer social skills and more behavior problems experienced more negative peer responses during SST sessions and less collateral support from teachers and parents. Hierarchical regressions (controlling for pretreatment scores) revealed attenuated intervention benefits for children who received more negative peer responses during SST and less parent and teacher support. The findings suggest that, within a sample of peer-rejected children, those with more severe skill deficits and elevated behavior problems at pretreatment have lower-quality SST experiences that reduce intervention benefits, even when normative classmates serve as peer partners. These issues warrant careful attention in future SST intervention design and research.