Entry into clinical trials of treatment for deep venous thrombosis (DVT) involves a screening process that excludes many patients with this condition. Therefore, to obtain a profile of patients with DVT that reflects actual day-to-day clinical practice, we initiated a prospective registry of 150 consecutive patients with DVT at Brigham and Women's Hospital, Boston. We reviewed the medical records of all patients who received a diagnosis of DVT from November 1, 1994 through March 31, 1995 and did not exclude any patients. Of the 150 patients, 120 (80%) were symptomatic and 30 (20%) were asymptomatic. Frequent baseline characteristics included surgery within the preceding 6 months (47%), cancer (26%), and previous venous thromboembolism (23%). DVT was diagnosed by ultrasonography (n = 145), contrast venography (n = 3), or clinical evaluation (in 2 patients who had technically unsatisfactory ultrasound results). Overall, 133 (89%) patients received anticoagulation; 5 (3%) received thrombolysis; and 27 (18%) underwent placement of an inferior vena caval filter. In 3 patients major hemorrhage developed from anticoagulation or thrombolysis. The 3-month mortality rate was 19%, including 10 patients who died during the index hospitalization. In no case was pulmonary embolism thought to be the cause of death. Ten patients had recurrent venous thrombosis during the initial 3 months after diagnosis. The most surprising findings in the DVT registry were the high rates of cancer and surgery, the high rate of filter placement, and the high mortality rate. These registry results differ substantially from findings in randomized controlled trials of DVT treatment.