Aspirin is one of the oldest drugs and has been purported to have multiple beneficial effects, including prevention of cardiovascular disease and cancer, in addition to its original indication for treatment of inflammation, fever and pain. In cancer chemoprevention studies using animal models, two methods of aspirin administration have been employed: oral gavage and diet. The untested assumption was that exposure and the resultant pharmacological effects are similar for these two administration methods when dosing is normalized on the basis of mg/kg body weight/day. This study examined and compared time-dependent plasma and colon mucosal concentrations of aspirin metabolite salicylate (aspirin concentrations were below level of quantification), plasma thromboxane B(2) concentrations, and colon mucosal prostaglandin E(2) concentration following these two different dosing paradigms in rats. Diet dosing yielded relatively constant plasma and colon salicylate concentration vs. time profiles. On the other hand, oral gavage dosing led to a rapid peak followed by a fast decline in salicylate concentration in both plasma and colon. Nevertheless, the exposure as measured by the area under plasma or colon concentration-time curve of salicylate was linearly related to dose irrespective of the dosing method. Linear relationships were also observed between colon and plasma salicylate areas under the curve and between colon prostaglandin E(2) and plasma thromboxane B(2) areas under the curve. Therefore, more easily accessible plasma salicylate and thromboxane B(2) concentrations were representative of the salicylate exposure and prostaglandin E(2) pharmacodynamic biomarker in the target colon, respectively.