BACKGROUND According to the Rotterdam 2003 criteria, an ovary is defined as polycystic if 12 or more follicles of 2–9 mm are present, when evaluating the ovary by ultrasonography on Days 3–5 of the menstrual cycle in women not using hormonal contraceptives. The aim of this population-based study was to estimate the prevalence of polycystic ovaries (PCO) in a representative sample of young Danish women according to the Rotterdam criteria. METHODS From a Danish pregnancy cohort established in 1988–1989, 267 (61%) young adult daughters agreed to participate in a clinical examination and 174 (40%) consented to vaginal ultrasound. Sufficient image quality in at least one ovary was obtained from 154 women. Both users and non-users of hormonal contraceptives were included and the examination was not restricted to a particular time of the menstrual cycle. RESULTS The median (range) age was 20.1 (19.5–21.0) years. The median follicle number per ovary was 14 (6–30) and 12 or more follicles were counted in 104 of the 154 women. Thus, the prevalence was estimated to 68% 95% confidence interval (CI): 60–74%. PCO were present in 80% (95% CI: 65–89%) of non-users (n = 44) of hormonal contraceptives. Of the 104 women with PCO, 41% (95% CI: 32–51%) could be defined as having polycystic ovary syndrome. CONCLUSIONS A very large proportion of the young women had PCO according to the Rotterdam 2003 criteria. As the number of follicles is higher at a younger age, we believe the Rotterdam criteria should be revised, particularly to avoid misdiagnosis in this age group.