Epidemiological research throughout the last 50 years has provided the long list of risk factors for venous thrombosis that are known today. Although this has advanced our current understanding about the aetiology of thrombosis, it does not give us all the answers: many people have several of these risk factors but never develop thrombosis; others suffer from thrombosis but have none. In this review, we discuss how risk factors for venous thrombosis can be interpreted with use of several epidemiological models. We comment on how to explain why risk factors for first venous thrombosis differ from recurrent venous thrombosis, and use a causal model to better understand risk of first and recurrent venous thrombosis.