The efficacy and tolerability profile of nimesulide was assessed in a double-blind, piroxicam-controlled trial in 49 patients with osteoarthritis of the knee joint. Nimesulide was administered orally as 100 mg twice daily, and piroxicam was administered as 20 mg orally in the morning and placebo in the evening. These patients were stratified into two groups: clinical assessment and magnetic resonance imaging evaluation of the knee. All patients were followed for 8 weeks with clinical assessment every 2 weeks; of these, 11 patients were selected for magnetic resonance imaging evaluation, and they were treated for 24 weeks. The principal efficacy parameters were improvement in osteoarthritis severity index, joint tenderness, swelling, and functional capacity of the patients. The final judgment of efficacy was made by the physician. Nimesulide improved all primary and secondary efficacy parameters with an activity comparable to piroxicam. Significant reduction in joint swelling and tenderness was observed as early as 2 weeks in the two treatment groups; however, a greater number of patients were relieved in nimesulide group. At 8 and 24 weeks, the number of patients with no joint swelling were 66.7% versus 50% and 80% versus 66.7% in the nimesulide and piroxicam groups (P <.05). Functional capacity at 8 weeks improved in 72.2% of nimesulide and 44.4% of piroxicam recipients. Mild adverse effects, mainly gastrointestinal, considered possibly related to treatment were recorded in 4 patients treated with nimesulide and in 12 patients treated with piroxicam. No significant change in the articulating cartilage of the tibiofemoral joint was observed over a 24-week period with either treatment. Nimesulide was as efficacious as piroxicam in reducing the inflammatory indices, and both drugs were equally well tolerated.