Objectives. We aimed to evaluate the effect of low-frequency electro- acupuncture (EA) in unmarried women with polycystic ovary syndrome (PCOS). Design. A total of 54 women with PCOS were randomly assigned to either the acupuncture group (n=27) or the sham acupuncture group (control, n=27) for a total of 32 treatments over 16 weeks. In the acupuncture group 26 patients and in the control group 20 patients completed the trial. Outcome Measures. Main measures were androgen levels including 17-alpha-hydroxyprogesterone (17-alpha-OHP), androstenedione (A2), testosterone (T) and dehydroepiandrosterone (DHEA) at 0, 24 and 48 hours after stimulation with a dose of human chorionic gonadotropin (HCG). Other measures included body mass index, waist-to-hip ratio, sex hormone levels, etc. Results. After treatment, there was no significant difference in the main measures between the 2 groups (P > .05), except for the DHEA level at 0 h of HCG stimulation (P= .024) and acne score (P <.05). Comparison within the acupuncture group found that 17-alpha-OHP and A2 levels at 0 h and DH EA levels at 24 h of HCG stimulation after treatment were significantly decreased (P < .05), whereas T levels at 24 h were significantly increased (P < .05). Comparison within the control group showed 17-alpha-OHP level at 0 h and 17-alpha-OHP and A2 and DHEA levels at 24 h after treatment were significantly lower (P < .05). In addition, weight, BMI, HCG and Ferriman-Gallwey score in the acupuncture group and LH/FSH ratio was significantly reduced in the control group. Conclusion. Traditional EA is slightly more effective than sham acupuncture in reducing DHEA secretion and the acne score. Meanwhile, sham acupuncture is not completely ineffective. The specific mechanism of the two may be different.