A population-based case-control study was conducted in men aged 60 or less to assess the risk of prostate cancer associated with vasectomy and other factors. Data were obtained from 216 case-control pairs by telephone interviews; this number represented 55% of all eligible cases. The matched pairs relative risk (RR) for vasectomy in ever married men was 1.4 with a 95% confidence interval (CI) of 0.9-2.3. There was a positive association between the number of years since vasectomy and prostate cancer risk (1-sided P = 0.01). Early age at first sexual intercourse was associated with increased prostate cancer risk (age less than 17 vs. 21+, RR = 2.3, 95% CI = 1.3, 4.0) but there were no consistent associations with number of sexual partners or frequency of sexual intercourse. Cigarette smoking was also associated with increased risk of prostate cancer (RR = 1.9, 95% CI = 1.2, 3.0) and there was a positive dose-response relationship with years of smoking (1-sided P = 0.001). We discuss the possible implication of the low response rate on each of these findings. To determine whether the association with vasectomy might have a hormonal basis, we compared levels of testosterone (T) and testosterone binding globulin-binding capacity (TeBG-bc) in 33 of the vasectomized control men with levels in 33 non-vasectomized controls of the same age, weight and height. T levels were higher in vasectomized than in non-vasectomized controls (1-sided P = 0.06). The ratio of T to TeBG-bc (an index of bioavailable T) was 13.5% higher in vasectomized men (1-sided P = 0.03).