The aim of this paper is to explain individual differences in lifestyle risk. Lifestyle risk has previously been identified as a key social mechanism which has strong direct effects on juvenile offending. Building on statements derived from the Situational Action Theory (SAT), we test the assumptions that (1) family structure explains individual differences in lifestyle risk, (2) these effects are moderated by mechanisms of social control, and (3) the effects of mechanisms of control are moderated by the effect of propensity to offend (morality and self-control). It is assumed that this model holds in different population segments such as subpopulations by gender and ethnic background. Results from two independently drawn urban samples in Belgium and Sweden are used to discuss the generalization of these findings. Mixed support is found for the `equality of effects' thesis