Objective:To verify whether the homeostatic model assessment (HOMA) test is a suitable method for the identification of metabolic deterioration in normal-weight patients affected by polycystic ovary syndrome (PCOS). Design:Prospective clinical study. Setting:Academic clinic and research environment in Cagliari,Italy. Patient(s):Forty-nine PCOS normal-weight adolescent subjects,and 50 eumenorrheic,normal-weight,nonhirsute controls matched for age and body mass index (BMI). Intervention(s):History and physical examination,oral glucose tolerance test (OGTT) and blood sampling,ultrasound. Main Outcome Measure(s):The HOMA score and integrated secretory area under the curve of insulin values (I-AUC) during the OGTT were calculated. Result(s):Normal insulin sensitivity was defined as upper control 95th percentile by HOMA values < 65.6,I-AUC at 180 minutes < 16,921,and I-AUC at 120 minutes < 11,817. When applying the calculated I-AUC cutoff,27 PCOS patients were classified as normoinsulinemic and 22 as hyperinsulinemic,whereas using the calculated HOMA cutoff,only 9 PCOS patients could be classified as insulin resistant (IR). Thirteen of the 40 non-IR PCOS patients presented with hyperinsulinemia; fasting glucose and insulin levels and HOMA scores were not sufficient to identify these subjects. Thus,the HOMA test displayed a low sensitivity (41%) and specificity (100%) in the diagnosis of the metabolic disorder disclosed by I-AUC.Moreover,analysis of I-AUC after 120 and 180 minutes revealed how the shorter evaluation period did not suffice for identification of all hyperinsulinemic subjects,implying an unrecognized condition in 11 of 22 subjects. Conclusion(s):In young,normal-weight patients with PCOS,the prevalence of hyperinsulinemia is not detectable by HOMA studies. The prevalence of IR was 18%according to HOMA evaluation,whereas hyperinsulinemia was found in 44%of subjects examined by I-AUC. Normal-weight,young PCOS patients should undergo a 3-hour OGTT to detect early metabolic abnormalities.