AIM: Mechanisms underlying the chemopreventive effects of cyclooxygenase (COX) inhibitors remain elusive. We have previously shown that celecoxib but not indomethacin could prevent carcinogen-induced gastric cancer development in Wistar rats. This chemopreventive effect appeared to be independent of COX-2 and prostaglandin (PG) E_2 suppression since the lowest PGE_2 was obtained in indomethacin group. This study compared the cell kinetic changes in stomachs of rats after treatment with celecoxib (5,10, 20 mg/(kg-d)) or indomethacin (3 mg/(kg·d)) to gain more insights into the chemopreventive mechanism. METHODS: The apoptosis and proliferation indexes in gastric tumor, adjacent non-cancer tissues and normal gastric tissues were determined. Apoptosis was quantified by apoptotic nuclei counting and TUNEL, whereas proliferation was determined by Ki67 immunostaining. RESULTS: Treatment with either celecoxib or indomethacin inhibited gastric tumor proliferation by more than 65% (P＜0.02). However, celecoxib caused a dose-dependent increase in apoptosis (P＜0.05) which was not seen in indomethacin-treated tumors (P=0.54). The highest apoptosis to proliferation ratio was seen in tumors treated with celecoxib at 10 mg/(kg·d). Treatment with this dose of celecoxib was associated with the lowest incidence of gastric cancer development. CONCLUSION: Our findings suggest that the difference in chemopreventive effects of indomethacin and celecoxib in this animal model of gastric carcinogenesis is largely due to the differential cell kinetic changes, which does not correlate with the degree of COX-2 and PG suppression. AIM: Mechanisms underlying the chemopreventive effects of cyclooxygenase (COX) inhibitors remain elusive. We have previously demonstrated that celecoxib but not indomethacin could prevent carcinogen-induced gastric cancer development in Wistar rats. This chemopreventive effect was to be independent of COX-2 and prostaglandin (PG) E_2 suppression since the lowest PGE_2 was obtained in indomethacin group. This study compared the cell kinetic changes in stomachs of rats after treatment with celecoxib (5,10, 20 mg / (kg-d)) or indomethacin (3 mg / METHODS: The apoptosis and proliferation indexes in gastric tumor, adjacent non-cancer tissues and normal gastric tissues were determined. Apoptosis was quantified by apoptotic nuclei counting and TUNEL, was determined by Ki67 immunostaining. RESULTS: Treatment with either celecoxib or indomethacin induced gastric tumor proliferation by more than 65% (P <0.02). However, celecoxib caused a dose-dependent increase in apoptosis (P <0.05) which was not seen in indomethacin-treated tumors (P = 0.54). The highest apoptosis to proliferation ratio was seen in tumors treated with celecoxib at 10 mg / (kg • d). Treatment with this dose of celecoxib was associated with the lowest incidence of gastric cancer development. CONCLUSION: Our findings suggest that the difference in chemopreventive effects of indomethacin and celecoxib in this animal model of gastric carcinogenesis is largely due to the differential cell kinetic changes, which does not correlate with the degree of COX-2 and PG suppression.