Data derived from a cohort of 50,897 pregnancies that occurred between 1959 and 1965 are currently being evaluated to determine whether any associations between maternal drug exposure and abnormalities in the offspring exist. In this report, the cohort being studied is briefly described, and the methods used to identify associations, together with factors that confound them, are outlined. Also presented are the techniques used to control for these factors in evaluating the associations. Issues concerning causal interpretation of associations that remain after confounding factors have been controlled are discussed.