Associated ultrasonic Doppler study (D) and plethysmography (P) were evaluated as compared to phlebography in diagnosis of venous thrombosis of the lower limbs (TVP). Probes (5 and 7.5 MHz respectively) were used in D for examination of proximal and calf veins. Plethysmography using a mercury ring gauge (Perivein, ETNA) was considered pathologic if filling volume was less than 1.5 vol% and/or emptying volume less than 40 ml/min/100 ml. One hundred and four patients hospitalized in an internal medicine department for suspected deep venous thrombosis were studied, including 97 for whom phlebography interpretation was unmistakable: 11 calf vein, 28 collector trunk and 42 total limb involvement TVP. In these localizations, D sensitivity was respectively 55, 72 and 100% (mean 85.5%) and that of P 63.5, 68 and 88% (mean 77.7%); association of D or P improved sensitivity: 82, 82 and 100% (mean 91.5%). Specificity was 86.6% for D, a poor 62.5% for P and 93.5% for associated D and P. The results are comparable to those in the literature but disappointing for diagnosis of isolated calf vein TVP. False-negatives in both D and P were due to strong collateral circulation or limited thrombus extension.