Objective: To test the hypothesis that oligo-ovulatory ovulation is associated with hyperandrogenemia in the absence of other clinical signs of androgen excess. Design: Retrospective cohort controlled. Setting: Tertiary care university-based outpatient clinic. Patient(s): Thirty-two consecutive nonhirsute oligo-ovulatory women and 37 healthy eumenorrheic controls. Intervention(s): All study subjects underwent a physical exam and blood sampling in the follicular phase of the menstrual cycle. Main Outcome Measure(s): Serum was assayed for total T, sex hormone-binding globulin (SHBG), androstenedione (A), and DS, and the free T level was calculated. Result(s): Mean free T was higher and SHBG was lower in oligo-ovulatory patients compared with controls. The oligo-ovulatory patients were subdivided according to severity of menstrual abnormalities into those with cycles < 45 days (n = 19, oligo-ovulatory-short) and ≤45 days (n = 13, oligo-ovulatory- long). Oligo-ovulatory-long and -short patients had similar mean DS, A, and total T levels; but mean free T was higher and SHBG was lower in oligo- ovulatory-long patients compared with either oligo-ovulatory-short women or controls. Five of 13 (38%) oligo-ovulatory-long patients had at least one abnormal androgen value compared with 1 of 19 (5%) oligo-ovulatory-short patients, which was a significant difference. Conclusion(s): Approximately 40% of nonhirsute oligo-ovulatory women with cycle intervals ≤ 45 days demonstrate at least one abnormally elevated androgen level, suggesting that they may have a discrete form of the polycystic ovary syndrome. Furthermore, this data brings into question the sensitivity of hirsutism as a marker for hyperandrogenemia.