We have compared the analgesic and opioid sparing effect of three i.v. non-steroidal anti-inflammatory drugs with placebo in a randomized, double-blind, placebo-controlled study in 80 adult patients after elective tonsillectomy. A standard anaesthetic was used. After induction of anaesthesia, patients received ketoprofen 100 mg, diclofenac 75 mg or ketorolac 30 mg by i.v. infusion over 30 min. Patients in the placebo group received saline. Ketoprofen and diclofenac infusions were repeated after 12 h and ketorolac infusion at 6 h and 12 h. Oxycodone was used as rescue analgesic. Patients in the ketoprofen group requested 32% less opioid and patients in the diclofenac and ketorolac groups 42% less opioid than those in the placebo group (P < 0.05). There were one, two and six patients in the placebo, diclofenac and ketorolac groups, respectively, but none in the ketoprofen group, who did not request opioid analgesia during the study (P < 0.05, ketorolac vs placebo and ketoprofen). Visual analogue pain scores were similar in all groups. Visual analogue satisfaction scores were significantly higher in the diclofenac group compared with the placebo group. The incidence of nausea was 44-54%. There were no differences in the incidence of other adverse reactions. We conclude that all three non-steroidal anti-inflammatory drugs were superior to placebo after tonsillectomy.