Conduct problems seem to increase among youths placed in group homes. An overview, however, suggested that group home resources may be protective. Smaller homes with fewer residents, may better prevent delinquency and criminality. It was theorized that the fewer the number of youths in a group home the more resourceful it is in terms of its youths/caregiver ratio. Group home size probably also matters because the more the concentration of residents increases the more their influence, pro or con, increases. Logistic regressions modeled conduct problems and tested the hypothesis that group home size moderates peer influence-conduct problem relationships such that large homes with many residents are relatively risky places, while smaller homes with fewer residents are relatively protected places. It was tested and supported cross-sectionally among 875 youths, 10 to 17 years of age in Ontario group homes in 2012–13 (participation = 90%) and longitudinally among 175 youths who remained in group home care three years later (2015–16, follow-up = 96%). The positive peer influence by group home size interaction at baseline was such that the influence of such peers was most protective in large group homes with eight or more residents. A main protective association of positive peers was maintained at follow-up, but not its interaction with group home size. The negative peer influence by group home size interaction was significant at cohort baseline and at 3-year follow-up. No youth ought to be placed in a group home with more than six or seven peers. Prospective cohorts are needed to definitively test this theory about the protective effects of relatively small group homes.