In ten large, well-controlled, randomised trials (n = 203 to 1089), valdecoxib, a selective inhibitor of cyclo-oxygenase-2, was significantly more effective than placebo in the treatment of osteoarthritis, rheumatoid arthritis and pain associated with primary dysmenorrhoea, and for postsurgical analgesia. Valdecoxib 1.25 to 10mg twice daily and valdecoxib 10mg once daily were more effective than placebo for the relief of pain in patients with osteoarthritis of the knee, and dosages above 5mg twice daily were similar in efficacy to naproxen 500mg twice daily. Similarly, valdecoxib 5 and 10 mg/day were as effective for osteoarthritis of the hip as naproxen 500mg twice daily. In patients with rheumatoid arthritis, valdecoxib 10, 20 or 40 mg/day was significantly more effective than placebo, and similar in efficacy to naproxen 500mg twice daily; there were no significant differences in efficacy between the three dosages of valdecoxib. Valdecoxib 20 or 40mg administered 1 to 3 hours before and 12, 24 and 36 hours after hip arthroplasty provided significantly better analgesia than placebo, and significantly reduced the amount of morphine taken by patients. Single doses of valdecoxib 10 to 80mg administered before foot or oral surgery provided significantly better analgesia than placebo; when administered after oral surgery, valdecoxib 20 or 40mg provided greater sustained analgesia than oxycodone 10mg/paracetamol 1000mg or rofecoxib 50mg. In contrast to three nonselective nonsteroidal anti-inflammatory drugs (NSAIDs), valdecoxib 40mg twice daily did not cause significant changes in platelet function and bleeding times. Chronic users of NSAIDs who were switched to valdecoxib 10 or 20 mg/day for 12 weeks experienced significantly fewer gastroduodenal erosions or ulcers than patients receiving ibuprofen 2400 mg/day or diclofenac 150 mg/day for 12 weeks. Valdecoxib was generally well tolerated in clinical trials, with a similar incidence of adverse events to placebo.