Since the clinical diagnosis of deep vein thrombosis (DVT) of the lower limbs is nonspecific, a confirmation by objective tests is mandatory. Recently, three new simple and reproducible methods of detecting DVT have been developed: computerized impedance plethysmography, standardized Doppler ultrasound, and compression ultrasonography. In three large prospective studies including consecutive outpatients with clinically suspected DVT, these tests were blindly evaluated versus phlebography, to determine diagnostic criteria and accuracy. The sensitivity for proximal DVT was 91% for computerized impedance plethysmography, 91% for standardized Doppler ultrasound and 100% for compression ultrasonography; sensitivity for all thrombi (including calf-vein thrombi) was 86, 85 and 91%, respectively; the specificity was 94% for computerized impedance plethysmography, and 99% for both standardized Doppler ultrasound and compression ultrasonography. The results of these studies demonstrate that all the three tests are highly specific and sensitive methods for the diagnosis of proximal DVT in symptomatic outpatients. However, isolated calf thrombi could not be detected adequately. Before these simple tests can be recommended as substitutes for phlebography, the safety of withholding anticoagulant therapy in patients with repeated normal tests should be assessed.