Many diseases are believed to occur as a result of interactions between genes and environment. In epidemiology, there are several study designs, such as cohort and case-control studies, that can be used to analyze biological interaction. In this chapter, we discuss crucial issues in the design stage of epidemiologic studies, how different measures of disease occurrence and measures of associations are calculated, and how systematic and random errors might hamper the accuracy of the results. By properly addressing these issues, one can design accurate epidemiologic studies and thus also decent studies of interaction. Later in this chapter, we give an example of how alcohol consumption and the genetic risk factor HLA-DRB1 shared epitope alleles interact in the development of rheumatoid arthritis, by using data from a Swedish population-based case-control study. The example demonstrates how interaction analyses in practice might be performed within epidemiologic study designs.