A consecutive series of 51 patients with suspected pancreatic cancer were studied in order to evaluate the contribution of cytologic data in establishing a definitive diagnosis of the tumor. A percutaneous ultrasonically-guided, and a peroperative fine-needle, aspiration were performed in 26 and 29 patients, respectively; 4 cases underwent both methods. Final histologic diagnosis in surgical or autopsy specimens showed cancer of the pancreas in 41 cases and chronic pancreatitis in 10. Cytology was successful in 33/41 cancer patients (80.5%): percutaneous aspiration was positive in 15/20 (75%) cases, and peroperative aspiration in 18/25 (72%). The difference between the two methods was not significant. Neither technique gave false positives in chronic pancreatitis. In none of the 51 patients was relevant morbidity reported. These cytological studies are at present accurate and mostly safer than histologic examinations. In particular ultrasound-guided percutaneous aspiration should be given precedence, since it provides preoperative confirmation of the tumor, thus obviating the need for exploratory laparotomy.