A review of the literature concerning the effect of environmental factors on crime and delinquency indicates that two types of population density (building density and household density) predict juvenile delinquency. Explanations which focus on both informal (built environment) and formal (police) mechanisms of social control have been developed. From a sample survey of 835 adolescents in a suburb of one of Canada's largest cities, this study eliminates the incongruity between such explanations and evidence by using the individual as the unit of analysis rather than official data. A path model is presented in which building density, household density, and socioeconomic status are exogenous variables; self-reported delinquency, marijuana use, and police presence are intervening variables; and police contact is the dependent variable. The major contribution of this study is the provision of empirical support for the idea that the built environment may affect both the informal and formal control of juveniles. Although building density does not predict illegal activities, it does predict marijuana use; this is consistent with the argument that the built environment may impede informal social control. The study also shows that building density design is a significant prediction of police presence, giving empirical support to the idea that high-density housing attracts formal agents of control, independent of the apparent level of adolescent deviant behavior. The relationship between informal and formal control and the built environment is complex: it is possible that building density both impedes informal control and elicits formal control. Further research is recommended. Tables, figures, and approximately 45 references are provided. (Author abstract modified)