Epidemiologic principles and study methods necessary in occupational health studies to make valid assessments of possible associations between occupational exposure and disease risk were discussed. Three basic epidemiologic study strategies, the cohort study, the case/control study, and the cross sectional study, distinguished by the temporal sequence of data acquisition, were described. Cohort studies evaluated the frequency of disease occurrence among a study population exposed to a certain risk factor. Cohort studies could be either retrospective or prospective, but retrospective studies were the most common form of cohort study in occupational epidemiology. In case/control studies, the retrospective progression from disease occurrence to exposure characteristics were investigated. Cross sectional studies consisted of the simultaneous characterization of exposure and disease within a population at risk. The strengths and weaknesses of these three study strategies were discussed. The indices of disease frequency employed in epidemiologic studies were defined. Sources of random and nonrandom error in epidemiologic studies and methods for controlling for potential confounders were described. The criteria for inferring causality in epidemiologic studies were outlined. The author concludes that confidence in making a causal inference should be based on the ability to consider many alternatives, any of which may be rejected.