BACKGROUND Adding a regional block to general anaesthesia can prevent postoperative pain and improve peripheral circulation. OBJECTIVE To seek improved postoperative analgesia and care due to a long-acting combined femoral and sciatic nerve block in patients undergoing femoropopliteal bypass surgery. DESIGN A randomised, double-blind, controlled trial. SETTING Vascular surgery unit of a French university hospital. PATIENTS Forty-four adults scheduled for bypass surgery under general anaesthesia. INTERVENTION Patients were allocated to receive either an active nerve block with 20 ml of 0.375% levobupivacaine and clonidine 0.5 μg kg –1 , or a simulated (sham) block only, but with local anaesthesia of the skin, before general anaesthesia. General anaesthesia was standardised with propofol, then sevoflurane and sufentanil adjusted according to clinical need. Postoperative analgesia was standardised with paracetamol 1 g every 6 h, and intravenous morphine, initially titrated in the postanaesthesia care unit and then patient-controlled. Oral analgesics were repeated up to day 3. MAIN OUTCOME MEASURES The primary outcome was morphine consumption during the first 24 postoperative hours. In a subgroup of postoperative patients distal tissue oxygen saturation was recorded at the lateral side of the blocked calf. RESULTS Patients in the active group received less intra-operative sufentanil (median dose 25 vs. 41 μg), needed less morphine during the first 24 h (15 vs. 27 mg) and 72 (20 vs. 35 mg) postoperative hours, than in the control group. They also had less pain on movement, but pain at rest, the tissue oxygen saturation and other rehabilitation outcomes were unaffected by the treatment. Tolerance outcomes were also similar between groups. CONCLUSION Combining the two regional blocks improves the quality of postoperative care in this frail population, probably by reducing the amount of peri-operative opioid. TRIAL REGISTRATION ClinicalTrials.gov (ref. NCT01785693).