This study examined 191 patients with 'reversed' and 99 patients with 'in situ' femoro-popliteal bypass technique. There were 85 diabetic patients (44.5%) in the group with 'reversed' bypass, and 43 patients (43.43%) in the 'in situ' group. There were 152 (79.68%) smokers in the 'reversed' bypass group, and 80 (80.8%) in the 'in situ' group. The graft patency was confirmed immediately after operation using CW Doppler and then followed up after 1, 6, l2 months and annually thereafter. The statistical analysis was performed using Pearsons chi-square test, Fischer's test and 'Life table' statistic methods. The patients were followed from 3 to 10 yr after surgery. 'In situ' bypass showed better patency than the 'reversed' bypass technique but only in the second and tenth follow-up year (P < 0.05). Also, 'in situ' bypass proved to be better than 'reversed' only in patients with one patent crural artery (P < 0.01). Diabetes and preoperative smoking did not significantly affect late patency regarding this technique (P > 0.05). However, continuous smoking after the operation significantly decreased late patency rate in both groups of patients (P < 0.01). There was no significant difference in the early thrombectomy rate between groups with 'reversed' and 'in situ' bypasses (P > 0.05). The early thrombectomy, however, significantly reduced late patency rate in both groups (P < 0.01). Therefore we suggest 'in situ' bypass in cases with poor run off, small-calibre vein and 'long' bypass. Also, we consider important more frequent physical and Doppler ultrasonographic control in patients who had early thrombectomy.