Compression ultrasound (US) is an excellent means of evaluating the femoral and popliteal veins but is generally regarded as inadequate for the diagnosis of calf vein thrombosis. This prospective study evaluated compression sonography of the calf veins in 45 symptomatic patients with normal femoral and popliteal veins. All patients underwent correlative venography. Compression US enabled identification of 15 of 17 patients with calf vein thrombosis (sensitivity, 88%). The two false-negative results were in patients with small isolated thrombi. Compression US results were true-negative in 26 of 27 patients with normal venograms (specificity, 96%). If these results can be duplicated by other investigators in larger series of patients, compression US will be an adequate screening modality for calf vein thrombosis.