Abstract Background Polycystic ovarian syndrome (PCOS) accounts for most cases of hyperandrogenism in adolescent females. Rarely, evaluation reveals testosterone concentrations (> 200 ng/dL). In those instances, imaging is recommended to evaluate for an ovarian or adrenal neoplasm. Recommendations are lacking for when imaging is nondiagnostic, yet testosterone levels continue to rise on standard treatment. Ovarian venous sampling may be considered as a diagnostic tool to locate the source of hyperandrogenism. A 12-year-old female was referred for evaluation of secondary amenorrhea, weight gain, and possible PCOS. Thelarche was at age 10 years and menarche at age 11 years. Eighteen months post-menarche she developed amenorrhea, acne, and hirsutism. She had no clitoromegaly. Her height was 50th percentile, weight >99 th percentile, and BMI 130th of the 95th percentile. She had no family history of PCOS. Initial lab evaluation showed mild hyperandrogenism: total testosterone of 52ng/dL (17-75), free testosterone of 1.3ng/dL (<0. 04-0.84), and DHEA of 423ng/dL (<202). OGTT done with only glucoses was normal. HbA1c was 5.6%. DHEAS, androstenedione, 17-OH progesterone, prolactin and TSH were normal. LH was 3.5mIU/mL, FSH 4.8mIU/mL, and estradiol 45pg/mL. Pelvic ultrasound showed enlarged right (25mL) and left ovaries (15.6mL). The follicular distribution was normal and bilateral simple cysts were noted. PCOS was suspected and she started on metformin and oral contraceptive pills. Her testosterone levels increased during subsequent visits with a peak total and free testosterone of 344ng/dL and 7.83ng/dL. HbA1c rose to 5.9%. She had fasting and postprandial insulin levels of 74.3 and 260mcIU/mL, respectively. Further evaluation revealed a normal AMH level of 6.4ng/mL (0.49-6.9), SHBG of 22.9nmol/L (7.7-108), a negative screen for anabolic steroid use, and 46,XX karyotype. Imaging over 8 months (multiple transabdominal ultrasounds, MRI of abdomen/pelvis and brain) was negative for neoplasm. The potential next steps discussed with the patient were continued serial imaging, GnRH agonist, or ovarian venous sampling with possible oophorectomy. Due to concern for neoplasm, the patient opted for venous sampling which showed a 10-fold gradient in total testosterone between right and left gonadal veins (2300ng/dL vs. 203ng/dL). She underwent right oophorectomy and switched from OCP to transdermal hormone therapy. Three months later, her total testosterone was normal (26ng/dL), she had lost 5kg and menstrual cycles returned. Pathology resulted as PCOS, with no evidence of neoplasm. Conclusions Localization of the source of severe hyperandrogenism can be a challenge. In our case, ovarian venous sampling results prompted unilateral oophorectomy, yet pathology showed no ovarian neoplasm. While ovarian venous sampling is a potential diagnostic tool, there is no data guiding when it should be used and interpretation of results, specifically if/when oophorectomy should be pursued. This is particularly relevant in the management of young, premenopausal females. Presentation: No date and time listed