A synergy between ketamine and methadone (ME) to produce antinociception has been demonstrated in experimental neuropathy. We wanted to compare post-operative opioid requirements in patients undergoing multilevel lumbar arthrodesis after the administration combined ME-ketamine (MK) or ME alone. This was a randomised double-blind study. During sevoflurane-remifentanil anaesthesia, 11 patients in each group received the following: ketamine bolus (0.5 mg/kg) after tracheal intubation, followed by an infusion of 2.5 μg/kg/min in the MK or saline bolus plus infusion in the ME group. Post-operative analgesia - during 48 h - was provided by patient-controlled analgesia (PCA), delivering bolus containing the following: ME 0.25 mg plus ketamine 0.5 mg in the MK group or ME 0.5 mg in the ME group. Lockout was 10 min, maximum of 3 boluses/h in both groups. Before closing the wound, all the patients received intravenous (i.v.) ME 0.1 mg/kg, dexketoprophen and paracetamol. Pain intensity was evaluated by a numerical rating scale (NRS), on arrival at recovery room (RR) and 24 and 48 h after surgery. In the RR, i.v. ME was administered until NRS was 3 when PCA was started. Dexketoprophen and paracetamol were administered 48 h. Remifentanil requirements were higher in the MK group (P = 0.004). Patients in the MK group received 70% less ME by PCA at 24 h (MK vs. ME group, median and interquartile range) - 3.43 mg (1.9-6.5) vs. 15 mg (9.65-17.38) (P < 0.001) - and at 48 h - 2 mg (0.5-3.63) vs. 9.5 mg (3.5-13.75) (P = 0.001). Patients in the MK group also attempted less doses, at 24 h: 19.5 (12.75-79.5) vs. 98 (41.5-137) (P = 0.043). Both groups had similar NRS values and comparable side effects. Perioperative ketamine-ME combination significantly decreased opioid consumption by PCA.