The overall purpose of this study is to contribute to bridging the gap between people- and place-oriented approaches in the study of crime causation. To achieve this we will explore some core hypotheses derived from Situational Action Theory about what makes young people crime prone and makes places criminogenic, and about the interaction between crime propensity and criminogenic exposure predicting crime events. We will also calculate the expected reduction in aggregate levels of crime that will occur as a result of successful interventions targeting crime propensity and criminogenic exposure. To test the hypotheses we will utilize a unique set of space-time budget, small area community survey, land-use and interviewer-led questionnaire data from the prospective longitudinal Peterborough Adolescent and Young Adult Development Study (PADS+) and an artificial neural network approach to modelling. The results show that people's crime propensity (based on their personal morals and abilities to exercise self-control) has the bulk of predictive power, but also that including criminogenic exposure (being unsupervised with peers and engaged in unstructured activities in residential areas of poor collective efficacy or commercial centres) demonstrates a substantial increase in predictive power (in addition to crime propensity). Moreover, the results show that the probability of crime is strongest when a crime-prone person is in a criminogenic setting and, crucially, that the higher a person's crime propensity the more vulnerable he or she is to influences of criminogenic exposure. Finally, the findings suggest that a reduction in people's crime propensity has a much bigger impact on their crime involvement than a reduction in their exposure to criminogenic settings.