Placebo and three doses of methylphenidate (MPH) were crossed with 3 levels of behavioral modification (no behavioral modification, NBM; low-intensity behavioral modification, LBM; and high-intensity behavior modification, HBM) in the context of a summer treatment program (STP). Participants were 48 children with ADHD, aged 5-12. Behavior was examined in a variety of social settings (sports activities, art class, lunch) that are typical of elementary school, neighborhood, and after-school settings. Children received each behavioral condition for 3 weeks, order counterbalanced across groups. Children concurrently received in random order placebo, 0.15 mg/kg/dose, 0.3 mg/kg/dose, or 0.6 mg/kg/dose MPH, 3 times daily with dose manipulated on a daily basis in random order for each child. Both behavioral and medication treatments produced highly significant and positive effects on children's behavior. The treatment modalities also interacted significantly. Whereas there was a linear dose-response curve for medication in NBM, the dose-response curves flattened considerably in LBM and HBM. Behavior modification produced effects as large as moderate doses, and on some measures, high doses of medication. These results replicate and extend to social-recreational settings previously reported results in a classroom setting from the same sample (Fabiano et al., School Psychology Review, 36, 195-216, 2007). Results illustrate the importance of taking dosage/intensity into account when evaluating combined treatments; there were no benefits of combined treatments when the dosage of either treatment was high but combination of the low-dose treatments produced substantial incremental improvement over unimodal treatment.