Background: Imaging often detects lesions suggestive of malignancy which need further diagnostic clarification. Methods: 527 fine-needle aspiration cytologies (FNAC), guided by ultrasound, carried out from 1986 through 1992 on 432 consecutive patients with abdominal and thoracic masses suspicious of malignancy, were reviewed. 222 tests were confirmed by histology and 305 by clinical follow-up. Results: 89% (n=469) of the samples were sufficient for interpretation, 6.8% (n=36) were questionable by our standards, 4.2% (n=22) were non-diagnostic. 377 true-positive, 131 true-negative, 1 false-positive and 18 false-negative results were obtained, with 95.4% sensitivity, 99.2% specificity, 96.4% accuracy, 99.7% positive and 87.9% negative predictive values. 267 examinations were done in patients with an unknown malignancy, 260 in patients with known malignancies yielding a sensitivity of 94.5 and 96.5%, respectively. The sensitivity in 183 liver, 127 retroperitoneal, 142 other abdominal, and 75 thoracic examinations was 96.8, 95.8, 93.4 and 95.1%, respectively. Out of 110 primary diagnostic procedures in which a final histologic diagnosis was available, FNAC was able to determine histogenesis in 72 cases; in 30 cases a correct generic diagnosis was made. Major surgery was avoided in 78 cases, a basis for treatment was established in 48 cases by FNAC alone, and a new approach to further management was taken in 37 cases. In 452 abdominal examinations no complications occurred, in 75 thoracic examinations 5 cases of pneumothorax occurred, one of them requiring chest drainage. Conclusions: Provided that there is sufficient experience in performing the aspiration technique, in specimen procurement and cytologic interpretation, ultrasonically guided FNAC of abdominal or thoracic masses is a reliable, easy-to-perform method with low risk. It yields valuable diagnostic information which may render more invasive procedures unnecessary, thus improving the quality of life of the cancer patient.