We prospectively studied 78 consecutive patients with acute pulmonary embolism (PE) to determine the most appropriate workup study for searching for hidden cancer. After a careful physical examination, the following tests were performed: erythrocyte sedimentation rate (ESR), complete blood cell counts, biochemistry, carcinoembryonic antigen levels, chest radiograph, upper gastrointestinal endoscopy, and abdominal ultrasound. If a malignant lesion was suspected, further appropriate tests were performed. After hospital discharge, periodic follow-up was performed on all patients in our outpatient clinic. A malignant lesion was detected in 9 of 78 patients: in 7 of them, cancer was diagnosed during the hospital admission because of acute PE. All but one of these 7 patients were asymptomatic, except for PE symptoms. In three of them some abnormalities on physical examination led to the diagnosis of cancer; in the remaining three patients the diagnosis was suspected from abnormal results of blood tests. Cancer was detected several months after hospital discharge in two additional patients: an esophageal cancer was diagnosed 5 months later in one of the 23 patients who refused endoscopy; and a colonic carcinoma was detected 21 months after hospital discharge in a patient in whom colonoscopy was not performed at the time of hospital admission. When considered overall, cancer was more commonly found in patients with "idiopathic" PE as compared with patients with known risk factors for PE development (6 of 21 patients vs 3 of 51 patients; p < 0.05). On the other hand, one patient died because of massive recurrent PE after a biopsy sample was obtained because of a prostatic node. Gross hematuria had developed shortly after biopsy, and any attempt to increase heparin doses was followed by recurrent hematuria. According to our experience, any decision about procedures that potentially involve bleeding should be carefully individualized in patients with acute PE.