To investigate the functional significance of CYP11alpha microsatellite polymorphism (tttta)(n) (-528 base pairs) in patients with polycystic ovary syndrome.Follow-up study.Academic research center.Eighty patients and 90 controls.Body mass indices and waist-to-hip ratios were determined. Blood samples were obtained for DNA analysis and hormone measurements.CYP11alpha marker (tttta)(n) genotyping and serum total testosterone levels.All the women were assigned to one of two genotype groups: 216+ (for women who had at least one copy of high frequency allele 216 with four repeat units) or 216- (for women who did not have allele 216). Fifty-nine patients (73.75%) had genotype 216+; their mean (+/-SD) total testosterone level was 78.0 +/- 19.8 ng/dL. Twenty-one patients (26.25%) had genotype 216-; their mean (+/-SD) total testosterone level was 100.0 +/- 23.3 ng/dL. The difference in total testosterone levels was statistically significant. Seventy-eight controls (86.67%) had genotype 216+ and 12 controls (13.33%) had genotype 216-; the total testosterone levels of these two groups were similar (38.6 +/- 15.5 vs. 40.3 +/- 12.1 ng/dL). The difference in genotype distribution between the women with polycystic ovary syndrome and the controls (26.25% vs. 13.33% with genotype 216-) was statistically significant.CYP11alpha (tttta)(n) allelic variants were associated with both polycystic ovary syndrome and total testosterone levels in women with polycystic ovary syndrome, suggesting the existence of an epistasis phenomenon.