Between March 1991 and June 1992, 600 patients were treated with mono-, bifemoral or iliofemoral arterial graft revascularization for occlusions and/or aneurysms. The patients were divided into two groups: group A (n = 296) received a Gelseal Vascutek graft immersed for 15 min before implant in a solution containing 1 mg/ml rifampicin; group B (n = 304) received an untreated Gelseal Vascutek graft. Both groups received perioperative antibiotic treatment with cephalosporins. Clinical follow-up was performed at 1, 6, 12 and 24 months after surgery to exclude signs of graft infection. Statistical analysis (X(2)) of pre-, intra- and postoperative risk factors showed both groups to be well matched. Among 600 patients treated, the 2-year follow-up showed 12 cases of graft infection (2.0%): five in group A (1.7%) and seven in group B (2.3%) (P = n.s.). All cases of graft infection originated in the groin and Staphylococcus aureus was isolated in 50% of cases. Statistical analysis (Mann-Whitney U test) showed a significant prevalence of lymphatic complications and immediate redo surgery in patients with graft infection. Of the 12 cases with infection, one was lost to follow-up, three were treated with total graft removal, six with partial graft removal and two with conservative therapy: there were no deaths. In spite of the relatively limited series and follow-up, no statistically significant difference emerged from the clinical use of vascular grafts pretreated with antibiotics.