The reported high sensitivity and specificity of impedance plethysmography (IPG) in the diagnosis of proximal vein thrombosis were evaluated in a prospective cohort follow-up study, in which IPG was performed three times over a period of seven days in 243 consecutive outpatients with clinically suspected deep venous thrombosis (DVT). The test was abnormal in 112 patients (46%). The positive predictive value of an abnormal IPG for venography-proved DVT was 90%. One hundred thirty-one patients (54%) with repeatedly normal tests were considered not to have DVT, and anticoagulants were withheld. During six months of follow-up, completed in all patients with repeatedly normal IPG, no patient died of venous thromboembolism and no patient returned with clinically suspected pulmonary embolism. One patient (0.8%) returned after two months with recurrent leg symptoms, and venous thrombosis was documented (95% confidence limits, 0.02% to 4.21%). Another patient, who was nonsymptomatic, had an abnormal IPG at the three-month follow-up visit, and venography revealed venous thrombosis. Patients sent by general practitioners to a community hospital, with clinically suspected acute DVT, can be effectively managed by serial IPG alone.