Background Early studies showed that naloxone infusion decreases the incidence of morphine-related side effects from intravenous patient-controlled analgesia. This study aimed to determine whether naloxone preserved analgesia while minimizing side effects caused by intravenous tramadol administration.rnMethods Eighty patients undergoing general anesthesia for cervical vertebrae surgery were randomly divided into four groups. All patients received 1 mg/kg tramadol 30 minutes before the end of surgery, followed by a continuous infusion with 0.3 mgkg-1·h-1 tramadol with no naloxone (group I, n=20), 0.05 μg-kg-1·h-1 naloxone (group II, n=20), 0.1 μg·kg-1·h-1 naloxone (group III, n=20) and 0.2 μg·kg-1·h-1 naloxone (group IV, n=20). Visual analog scales (VAS) for pain during rest and cough, nausea five-point scale (NFPS) for nausea and vomiting, and ramsay sedation score (RSS) for sedation were assessed at 2, 6,12, 24 and 48 hours postoperatively. Analgesia and side effects were evaluated by blinded observers. Results Seventy-eight patients were included in this study. The intravenous tramadol administration provided the satisfied analgesia. There was no significant difference in either resting or coughing VAS scores among naloxone groups and control group. Compared with control group, sedation was less in groups II, III, and IV at 6, 12, and 24 hours (P <0.05); nausea was less in groups II, III and IV than group I at 2, 6, 12, 24 and 48 hours postoperatively (P <0.05). The incidence of vomiting in the control group was 35% vs. 10% for the highest dose naloxone group (group IV) (P<0.01). Conclusion A small-dose naloxone infusion could reduce tramadol induced side effects without reversing its analgesic effects.rn