Aims and background The prognosis for ovarian cancer patients depends on the stage at diagnosis. As a prerequisite for any proposed procedure for ovarian cancer screening, high levels of specificity should be obtained using combinations of tests. Based on preliminary data from an ongoing feasibility study, this report is focused on the possible gain in specificity provided by fine-needle aspiration of occult ovarian masses detected by ultrasound. Methods Ultrasonography was used as a basic test, and fine-needle aspiration was obtained from selected patients. Those with (a) positive aspiration cytology or histology, (b) complex or solid masses showing volume increase, (c) complex or irregular masses and inadequate samples, and (d) recurrent cystic lesions were operated on. Results A total of 3541 asymptomatic patients 50-69 years of age underwent ultrasonography, and 98 were selected for fine-needle aspiration. Of these, 19 (positivity rate 19/3541 or 0.5%) were operated on. Two ovarian cancer cases (FIGO Stage II and III) were detected (detection rate 2/3541 or 0.6 × 1,000; positive predictive value 2/19 or 10.5%). Specificity was 3443/3539 or 97.3% for ultrasound alone and 3522/3539 or 99.5% for the procedure as a whole. Surgical exploration for relapses of benign cysts reduced by some 50% the potential gain in positive predictive value provided by aspiration cytology compared with ultrasound alone. Conclusions Although fine-needle aspiration has improved the performance of ultrasound, this 2-levei procedure does not appear to achieve acceptable levels of specificity.