Epidemiological studies and clinical observations suggest that nonsteroidal anti-inflammatory drugs and certain selective cyclooxygenase (COX)-2 inhibitors may reduce the relative risk of clinically evident prostate cancer. This prompted us to investigate the chemopreventive potential of celecoxib, a selective COX-2 inhibitor, against prostate carcinogenesis in a transgenic adenocarcinoma of the mouse prostate (TRAMP) model. Similar to prostate cancer in humans, prostate malignancies in TRAMP mice progress from precursor intraepithelial lesions, to invasive carcinoma that metastasizes to lymph nodes, liver, lungs, and occasionally to bone. The basal enzyme activity and protein expression of COX-2 is significantly higher (>4-fold) in the dorsolateral prostate of TRAMP mice up to 24 weeks of age compared with their nontransgenic littermates. Eight-week-old TRAMP mice were randomly divided and fed either control diet (AIN 76A) or a custom prepared AIN 76A diet containing 1500-ppm celecoxib ad libitum for 24 weeks, a dosage that would compare with the normal recommended dose for the treatment of human disease. Studies from two independent experiments, each consisting of 10 mice on test, showed that the cumulative incidence of prostate cancer development at 32 weeks of age in animals fed with AIN 76A diet was 100% (20 of 20) as observed by tumor palpation, whereas 65% (13 of 20), 35% (7 of 20), and 20% (4 of 20) of the animals exhibited distant site metastases to lymph nodes, lungs, and liver. Celecoxib supplementation to TRAMP mice from 8-32 weeks of age exhibited significant reduction in tumor development (5 of 20) with no signs of metastasis. Celecoxib feeding resulted in a significant decrease in prostate (56%; P < 0.0003) and genitourinary weight (48%; P < 0.008). Sequential magnetic resonance imaging analysis of celecoxib-fed mice documented lower prostate volume compared with the AIN 76A-fed group. Histopathological examination of celecoxib-fed animals showed reduced proliferation, and down-modulation of COX-2 and prostaglandin E2 levels in the dorsolateral prostate and plasma, respectively. These results correlated with retention of antimetastasis markers, viz E-cadherin, and alpha- and beta-catenin, along with a significant decrease in vascular endothelial growth factor protein expression. Celecoxib supplementation also resulted in enhanced in vivo apoptosis in the prostate as monitored by several techniques including a recently perfected technique of 99mTc-labeled annexin V in live animals followed by phosphor imaging. One striking observation in an additional study was that celecoxib feeding to mice with established tumors (16 weeks of age) significantly improved their overall survival (P = 0.014), compared with AIN 76A-fed group. Our findings suggest that celecoxib may be useful in chemoprevention of prostate cancer.