The economic costs of chronic health conditions and severe illnesses like diabetes, coronary heart disease or cancer are immense. Several clinical trials give information about the importance of individual behaviour for the prevalence of these illnesses. Changes in health relevant behaviour may therefore lead to a decline of avoidable illnesses and related health care costs. In this context, we use German micro data to identify determinants of smoking, drinking and obesity. Our empirical approach allows for the simultaneity between adverse health behaviour and self-reported health as a measure of the individual health capital stock. We can show that health behaviour is related to the socioeconomic status of an individual. Furthermore, we find gender-specific differences in behaviour as well as differences in the determinants of drinking, smoking and heavy body weight in particular.