Building on the spectacular success of molecular genetics in defining the biological basis of many rare single gene disorders over the past decade, epidemiologists have turned their attention to unravelling the complex genetic mysteries of common disorders, such as rheumatoid arthritis (RA). As a prelude to any such endeavour it is obviously important to establish that there is a significant genetic component to the disease. The classical approaches of twin and other family recurrence risk studies, coupled with prevalence studies in different ethnic and migrant populations, have been used to estimate the environmental and genetic contributions to RA. However, developing a consensus on these estimates has proved difficult, thereby providing an early warning to the unwary investigator that the road to gene discovery in RA is likely to be a rough ride.