BACKGROUND/AIMS: Indomethacin inhibits both cylooxygenase-1 (COX-1) and COX-2, whereas meloxicam selectively inhibits COX-2. We evaluated the effect of indomethacin and meloxicam on the gastric mucosa and serum concentration of tumor necrosis factor (TNF)-alphain rats. METHODS: Eighty rats were administered with indomethacin and meloxicam. They were divided into 30 min and 3 hours groups according to the time interval between the administration and the evaluation of gastric mucosal damage and serum concentration of TNF-alpha. Each group (n=40) was equally subdivided into 5 groups: control, indomethacin 3 mg/kg and indomethacin 30 mg/kg, meloxicam 3 mg/kg and meloxicam 30 mg/kg. The gastric damage score was assessed by gross observation and the concentration of serum TNF-alphawas measured by enzyme linked immunosorbent assay. RESULTS: At 3 hours after the administration, the gastric mucosal damage more was significant in the indomethacin administered group of a dose (3 mg/kg or 30 mg/kg) than in the other groups (p<0.05). The meloxicam group of a dose of 30 mg/kg revealed a significant gastric mucosal damage, as compared to the control. Gastric mucosal damage was correlated well with the serum level of TNF-alpha(r=0.759: p<0.001). CONCLUSIONS: Considering gastric mucosal damage, selective COX-2 inhibitor, meloxicam, seems to be relatively safe at a low dose as compared to non-selective COX inhibitors.