PURPOSE: To compare the tourniquet technique (TT) with the conventional venogram (CV) and the carbon dioxide venogram (CO2V) for the evaluation of central vein patency of the upper extremity. MATERIALS AND METHODS: CO2V, TT, and CV were performed on 100 upper extremities prior to an arteriovenous fistula operation. The central vein was divided into four segments. The best image of the venograms for each segment was chosen as a reference, and the venogram techniques for each segment were graded from 1 (invisible) to 5 (excellent) compared with those of the reference image. The grades of the various venogram techniques at each segment of the vein were compared statistically. RESULTS: For the SVC segment, the mean grades of CO2V, TT, and CV were 4.32, 3.60, and 2.45, respectively. TT is statistically superior to CV but inferior to CO2V. On the brachiocephalic vein, the mean grades of CO2V, TT, and CV were 4.41, 4.37, and 2.77 and were 4.81, 4.85, and 3.78 for the subclavian vein and 4.75, 4.93, and 4.57, respectively, on the axillary vein. On these segments, TT was statistically superior to CV, but no difference was noted with CO2V. CONCLUSION: TT is superior to CV in every segment of the central vein and presents similar values to those of the CO2V, except for the SVC.