Previous epidemiological studies of cryptorchidism have led to the hypothesis that the risk of undescended testis is associated with excess oestrogen exposure during pregnancy. A case-control study was undertaken to test this hypothesis, comparing mothers of affected boys (244) and normal male births (488) born within six months of a case selected randomly from the British Columbia population. Information was collected on the mother's reproductive history, family history, and past medical history, and events surrounding all pregnancies ending in a birth. The results were analysed using both the population-based sample of male births and the male sibs of cases as control groups. Neither exogenous oestrogen exposure, nor any of the pregnancy-related variables hypothesized to be indirect indicators of endogenous oestrogen exposure, including bleeding and nausea and/or vomiting, were found to be significantly associated with risk of undescended testes in either comparison. More mothers with later index births reported menstrual irregularity greater than half the time, and smoking, thought to have a protective effect, was more prevalent among case mothers than control mothers. No other variables were significantly different between case and control mothers. The results of this study do not support the hypothesis that elevated exogenous or endogenous oestrogen exposure during pregnancy increases the risk of undescended testis in male children.