The linkage between potentially teratogenic factors was studied in a material of 599 children with oral clefts and their matched controls. A method based on Yule's Q coefficient describing the degree of association between two dichotomous variables was applied. All factors studied (five groups of drugs taken by the mothers during early pregnancy, maternal influenza and fever) were significantly associated with the birth of children with clefts. The only factor whose association with clefts was explained by linkage to other factors was fever. In addition, the association between clefts and antipyretic analgesics other than salicylates could be partly explained by controlling the intake of salicylates. Although there was a strong association between influenza and consumption of salicylates, the correlation of neither of the two factors with oral clefts could be even partly explained by controlling the other. The method is considered suitable for epidemiological studies of congenital defects.