Objectives: Outcomes of childhood obesity interventions are rarely reported beyond 1 year. We hypothesized that the impact on the BMI z score from a child-centered physical-activity program in combination with a parent-centered dietary-modification program would be greater than either program conducted alone at 24 months' after baseline. Participants and methods: A total of 165 overweight prepubertal children (68 boys, aged 5.5–9.9 years, mean BMI z score: 2.8) were randomly assigned to either a child-centered physical-activity program, a parent-centered dietary-modification program, or both conducted together in an assessor-blinded 6-month intervention. Results: Using linear mixed models, all groups reduced their mean (95% confidence interval) BMI z score at 24 months from baseline (P < .001) (the activity and diet group: −0.24 [−0.35 to −0.13]; the diet-only group: −0.35 [−0.48 to −0.22]; activity-only group −0.19 [−0.30 to −0.07]). There was a significant group-by-time interaction (P = .04) with the activity + diet and the diet-only groups showing a greater reduction than the activity-only group. For waist z score and waist-to-height ratio, there was a significant time effect (P < .0001) at 24 months but no between-group differences (P > .05). Some metabolic outcomes improved at 24 months, although there were no between-group differences (P > .05). Conclusions: A reduction in BMI z score was sustained at 24 months by treatment with either program combination. The greatest effects were achieved through inclusion of a parent-centered diet program, indicating the importance of targeting parents within treatment and the possibility of targeting them exclusively in treating obese prepubertal children.