The study compared the analgesic efficacy and safety of two preoperatively administered cyclo-oxygenase-2 inhibitors, celecoxib and rofecoxib. Ninety adult patients undergoing thyroid surgery were divided into three groups (each n = 30). They were given a single oral dose of placebo, celecoxib 200 mg or rofecoxib 50 mg 1 h before induction of anaesthesia. All patients received a standard anaesthetic. Intraoperative blood loss was measured. Pain scores, sedation scores, heart rate, mean arterial pressure and respiratory rate were noted at 0, 1, 2, 4, 6, 12 and 24 h postoperatively. Analgesic (meperidine) requirements and adverse effects were recorded during the first postoperative 24 h. Compared with placebo, pain scores were significantly lower with rofecoxib at all time points (P < 0.05) and were significantly lower with celecoxib (P < 0.05) during the first 4 h. Pain scores were significantly lower with rofecoxib compared with celecoxib at 6, 12 and 24 h (P < 0.05). The average cumulative 24 h meperidine dose was significantly lower with both celecoxib (54.9 +/- 34.4mg) and rofecoxib (42.8 +/- 40.9 mg) compared with placebo (76.8 +/- 6.2 mg) (P < 0.01 and P < 0.001, respectively). There were no differences in the intraoperative blood loss, heart rate, mean arterial pressure, respiratory rate, sedation scores and incidence of adverse effects among groups. The preoperative administration of rofecoxib 50 mg and less so of celecoxib 200 mg provide a significant analgesic benefit with regard to postoperative pain relief and decrease in additional opioid requirements after thyroid surgery.