Children with conduct problems and callous-unemotional (CU) traits are at risk for multiple problems. Outcome research and mediation analyses testing for mechanisms of change in CU traits have been limited. We examined whether parent training-in a short-term (Brief Parent Training; BPT) or a comprehensive format (Parent Management Training, Oregon Model; PMTO)-or child-directed social skills training (Individual Social Skills Training [ISST]) produced positive effects on CU traits. In mediation models we tested parenting practices as mechanisms of change for CU traits. We pooled data from three randomized effectiveness trials, and 551 families were included in this study. Families had children between 3 and 12 years of age and displayed emerging or present conduct problems at home, day care, or school (BPT M age = 7.28, 31.9% girls; PMTO M age = 8.56, 36.5% girls; ISST M age = 7.64, 19.7% girls). Assessments were completed preintervention, postintervention, and at follow-up (6 months following intervention). Both BPT (d = .32) and PMTO (d = .39) had positive effects on CU traits at posttest, whereas ISST did not (d = -.06). At follow-up, only PMTO produced a significant effect (d = .48) on CU traits. A significant indirect effect on CU traits emerged by positive parenting. Both parent training conditions outperformed ISST. Only PMTO maintained its effects at follow-up. The findings suggest that PMTO can reduce CU traits and that improved positive parenting is associated with positive outcomes for children's CU traits.