Life style research has become an independent branch of social structural analysis that systematically complements the role of classical research on social inequality in explaining social action. However, there are important gaps in life style research. First, theoretical bases for explaining the statistical connection between life styles and social action are lacking. Secondly, it remains unclear why life styles sometimes are good predictors of social action and sometimes are not. In a first step this article puts forward an action-theoretic framework as a basis for life style research that implies a conceptual shift from life styles to cultural preferences. In a second step it develops three hypotheses that set out the conditions for a high or low explanatory power of cultural preferences. They are empirically tested by way of a data analysis that supports the applicability of the theoretical ideas developed in the paper.