Cyclo-oxygenase-2 overexpression has been described in experimental colitis. However, there are controversial findings suggesting that its inhibition by selective cyclo-oxygenase-2 inhibitors not only may have a beneficial effect on experimental colitis, but also exacerbate the inflammation-associated colonic injury. Thus, the role of cyclo-oxygenase-2 inhibitors in the possible modulation of colon inflammation is controversial and remains uncertain. In this study, we evaluated the effects of the selective cyclo-oxygenase-2 inhibitor, rofecoxib, on the extent and severity of ulcerative colitis caused by intracolonic administration of 2,4,6-trinitrobenzene sulphonic acid (TNBS) in rats. The lesions and the inflammatory response were assessed by histology and measurement of myeloperoxidase activity. Interleukin-1 beta, prostaglandin E(2) and D(2) levels in colon mucosa and the immunohistochemical expression of the cyclo-oxygenases-1 and -2 were also studied. Finally, we investigated the effects of rofecoxib on apoptosis of colonocytes by the appearance of DNA fragmentation. Inflammation following TNBS was characterized by increased colonic wall thickness, oedema, diffuse inflammatory cell infiltration in the mucosa, and necrosis. Increased myeloperoxidase activity, as an index of neutrophil infiltration in the mucosa, and interleukin-1 beta levels were also measured in the colon. Administration of rofecoxib significantly (P<0.05) reduced the colonic damage, the degree of neutrophil infiltration, and interleukin-1 beta levels. In addition, apoptosis was significantly increased in TNBS-treated rats, but not in rofecoxib plus TNBS-treated rats. We concluded that rofecoxib seems to have beneficial effects in TNBS-induced colitis by diminishing the initial stage of inflammation, probably by a mechanism related to inhibition of prostaglandin E(2) by the cyclo-oxygenase-2 pathway. The data suggest that cyclo-oxygenase-2-selective inhibitors may have a therapeutic role in ulcerative colitis.