A prospective cooperative study was performed to assess the relative efficacy of computed tomography (CT) and ultrasound in detecting and identifying pancreatic lesions. Of the 279 patients in the study, 146 were found to have a normal pancreas, and 133 had an abnormal pancreas. All patients underwent both CT and ultrasound examinations. Forty-four ultrasound examinations were technically unsatisfactory. When these suboptimal examinations were excluded, CT had a sensitivity of 0.87 and a specificity of 0.90 in detecting an abnormal pancreas. Ultrasound had a sensitivity of 0.69 and a specificity of 0.82. In detecting a lesion and identifying it as malignant or inflammatory, CT had a sensitivity of 0.84 and ultrasound had a sensitivity of 0.56. It is concluded that CT is the method of choice for detecting a pancreatic lesion, assessing its extent, and defining its etiology.