The descriptive epidemiology of type 2 diabetes suggests that gene-lifestyle interactions are critical to the development of the condition. However, unravelling the molecular detail of these interactions is a complex task. The existing literature is based on small intervention studies or cross-sectional observational quantitative trait studies. Our systematic review of the literature identified some evidence of interactions, most notably for a common variant in the PPAR-gamma gene which appears to interact with the nature of dietary fat intake. Other interactions have been reported for adrenoceptors, uncoupling proteins, fatty acid binding proteins, apolipoproteins and lipoprotein lipase. There are, to date, no reports based on the ideal study design which is a case-control study nested within a cohort. To limit the likelihood of false discovery, such studies would need to be large and the search for interaction should be restricted to a priori biologically driven hypotheses. Additional study designs that examine differential response to lifestyle change or test interaction in the context of quantitative trait studies would complement the nested case-control approach, but the emphasis here should be on precision of measurement of both phenotype and lifestyle behaviour.