Background Behavioral parenting programs are a primary strategy used to reduce children's conduct problems. Although behavior problems in siblings may co‐occur, behavioral parenting program trials typically report outcomes for one child per family (the index child), with potential program effects on any non‐targeted sibling largely neglected. This study examined co‐occurring patterns of index child and non‐targeted sibling conduct problems, and how parental participation in the Incredible Years (IY) program changes these patterns. Methods We used individual participant data pooled across three randomized trials of the IY parenting program in England, Wales, and Ireland, with data for the index child and one non‐targeted sibling ( N = 240 families, 480 children; index child: M age = 4.73, SD = 1.44, range 2–9 years, 62% male; non‐targeted sibling: M age = 5.94 years, SD = 3.15, range 6 months−15 years, 49% male). We used latent transition analysis to identify latent classes at both baseline and posttest based on families' combinations of index child and non‐targeted sibling conduct problems. Results We identified two classes with distinct patterns of co‐occurring sibling dyad conduct problems: one with moderate clinical levels of index child conduct problems and non‐clinical levels for the non‐targeted sibling; and one with severe clinical levels for both children. In terms of the effects of IY, most intervention families maintained their patterns of sibling dyad conduct problems, but with lower levels across classes. Most intervention families reported improvements predominantly for the index child. However, a minority of families with severe baseline levels of conduct problems in both children moved to a class with non‐clinical levels for both children. Conclusions For most families, IY had limited effects on non‐targeted sibling disruptive behavior. However, IY may reduce co‐occurring sibling conduct problems for a small number of families with initially severe levels in both children.