A case-control study of in utero estrogen exposure and cryptorchism was carried out using as cases males born in Rochester, Minnesota, during the years 1943-1973 who were diagnosed as having cryptorchism. Two different control groups were selected for comparison, control group I being more closely matched than control group II. The estimated relative risks (RR) for estrogen exposure were 1.3 (95% confidence interval (CI) = 0.5-3.1) and 1.1 (95% CI = 0.5-2.6) for control groups I and II, respectively. In the univariate analysis, the only significantly elevated relative risks found were those for bleeding and spotting in the third trimester for cases versus control group II (RR = 3.7; 95% CI = 1.1-15.7), birth weight less than 2,500 g for cases versus control group II (RR = 3.4; 95% CI = 1.3-9.9), and gestational age of 40 weeks or less for cases versus control group I (RR = 1.8; 95% CI = 1.2-2.9). No elevated relative risks were associated with other problems during the index pregnancy or with prior pregnancies, nor with progestin exposure, smoking, presentation at delivery, or mode of onset of labor. Multivariate analysis also provided no evidence to suggest that in utero estrogen exposure is associated with cryptorchism in male offspring.