This study examines whether the Communities That Care (CTC) prevention system influences targeted risk and protective factors and the subsequent development of problem behaviour among adolescents (12-18 years) in the Netherlands. In this quasi-experimental study of ten communities (five experimental, five control), adolescent outcomes were examined using a longitudinal design. Propensity score methodology and multilevel modelling were used to test the effects. No impact of CTC could be detected on the development of a broader range of behaviours, or on the initiation of drinking and smoking. These results differ from earlier experimental studies and could be attributed to the threats to internal validity and design limitations of this community study. It is important to identify, monitor, address and report threats to internal validity when conducting effectiveness trials.