Recent chemopreventive studies in our laboratories showed that the COX-2 inhibitor, celecoxib, inhibited the induction of mammary cancer by 7,12-dimethylbenz(a)anthracene (DMBA). In this study, we examined the relative chemopreventive effect of varying doses of celecoxib, on the development and growth of DMBA-induced rat mammary tumors. At 10 days prior to receiving a single intragastric dose of 15 mg DMBA / rat, female Sprague-Dawley rats were fed a control chow diet or diets containing 250, 500, 1000 or 1500 ppm celecoxib until termination of the experiment. Administration of increasing doses of celecoxib inhibited mammary tumor incidence and multiplicity as well as tumor volume in a dose-dependent manner. At 122 days post DMBA-intubation, mammary tumor incidence was 100% in the control rats compared to 80%, 50%, 45% and 25% in rats receiving 250, 500, 1000 or 1500 ppm celecoxib, respectively (p<0.001). Similarly, tumor multiplicity and tumor volume were significantly reduced by increasing the dose of celecoxib from 250 to 1500 ppm in the diet. The control rats had an average of 3.46 tumors/rat compared to 1.80, 1.00, 0.75 and 0.50 tumors/rat in animals receiving 250, 500, 1000 or 1500 ppm celecoxib, respectively (p<0.001). Average tumor volumes in rats fed 250, 500 1000 or 1500 ppm celecoxib were 0.42,0.34,0.31 and 0.16 cm(3) compared to 1.29 cm(3) in the control rats(p<0.001). There was a concomitant increase in the steady-state serum concentration of celecoxib with the dose. These results indicate that, in this rat model, the chemopreventive effect of celecoxib against breast cancer is dose-dependent and that celecoxib is effective even at lower dose levels.