Repair of major venous injuries of the extremities has been advocated to improve limb salvage rates and to prevent the early and late sequelae of venous interruption. The contribution of venous repair to the surgical outcome remains controversial, however, in part because the fate of venous reconstruction has previously not been well defined. The current study was done to determine the early patency rate of venous repair, to compare the accuracy of various methods used to assess venous patency, and to analyze the relationship between early venous patency and surgical outcome. During a recent 27-month period, 36 patients with major extremity venous injuries were treated by venous reconstruction; 34 patients (94%) had an associated major arterial injury that also required repair. Venous repair was performed in the upper extremity (22%) as well as the lower extremity (78%) using various reconstructive methods, including lateral repair (17%), end-to-end anastomosis (11%), autogenous vein patching (25%), interposition autogenous vein grafting (42%), and panelled autogenous vein grafting (6%). After operation, venous repair patency was evaluated by clinical examination, impedence plethysmography, and Doppler ultrasonography, and contrast venography. There were no perioperative deaths in these 36 patients. The limb salvage rate was 100% and all 34 major arterial repairs were patent at the time of hospital discharge. Venography performed on the seventh postoperative day demonstrated that 14 venous repairs had thrombosed (39%) and that 22 had remained patent (61%). Local venous repair had a significantly lower thrombosis rate (21%) than those requiring interposition vein grafting (59%) (p less than 0.03). Compared with venography, the clinical evaluation was 67% accurate in the assessment of venous repair patency, and the noninvasive examination was 53% accurate. In conclusion, a substantial percentage of venous repairs will thrombose in the postoperative period, especially if interposition vein grafting is used. However, in this series limb salvage was not adversely influenced by an unexpectedly high rate of venous repair thrombosis. In addition, clinical evaluation and noninvasive testing did not provide an accurate assessment of venous patency after venous repair.