One hundred and two patients, presenting at the outpatient departments of two Boston teaching hospitals, underwent clinical examination and venography. History, physical examination and presence of risk factors were unreliable in the diagnosis of deep venous thrombosis. Seventy-two per cent of the patients with deep venous thrombosis had a proximal extension of the thrombus to the femoral vein or higher. The outpatient with deep venous thrombosis appears to differ from the more frequently studied inpatient in the time of diagnosis and its anatomic extent. The high incidence of false-positive clinical examination results has important cost-benefit implications.