A T-to-C polymorphism in the 5' promoter region of the CYP17 gene that encodes the cytochrome P450c17alpha has been implicated as a risk factor for prostate cancer, but individual studies have been inconclusive or controversial. Therefore we performed a meta-analysis of 10 studies (12 comparisons) with CYP17 genotyping on 2404 patients with prostate cancer and 2755 controls. Overall, the random effects odds ratio (OR) for the A2 (C) versus A1 (T) allele was 1.08 [95% confidence interval (CI), 0.95-1.22], with some between-study heterogeneity (P = 0.04). There was no suggestion of an overall effect either in recessive or dominant modeling of A2 effects, and the comparison of A2/A2 versus A1/A1 also showed no differential susceptibility to prostate cancer (OR, 1.15; 95% CI, 0.91-1.46). No effect of A2 was seen in subjects of European descent (7 comparisons, OR, 1.04; 95% CI, 0.92-1.18, no significant between-study heterogeneity) or Asian descent (2 comparisons, OR, 1.06; 95% CI, 0.66-1.71; P = 0.02 for heterogeneity), whereas A2 increased susceptibility to prostate cancer in subjects of African descent (3 comparisons, OR, 1.56; 95% CI, 1.07-2.28; no between-study heterogeneity). Smaller studies unilaterally showed more prominent genetic effects for A2 than larger studies (P = 0.038). The meta-analysis suggests that the CYP17 polymorphism is unlikely to increase considerably the risk of sporadic prostate cancer on a wide population basis, especially in subjects of European descent. Previously reported associations may reflect publication bias, although it is also possible that the polymorphism may be important in subjects of African descent.