1. The metabolism and pharmacokinetics of two structurally similar PPAR agonists, MRL-I, (2R)-7-[3-[2-chloro-4-(4-fluorophenoxy)phenoxy]propoxy]-2-ethyl-3,4-dihydro-2H-benzopyran-2-carboxylic acid, and MRL-II, (2R)-7-[3-[2-chloro-4-(2,2,2,-trifluoroethoxy)phenoxy]propoxy]-3,4-dihydro-2-methyl-2H-benzopyran-2-carboxylic acid, in dogs were investigated. 2. MRL-I was absorbed rapidly in dogs and exhibited linear pharmacokinetics over the dose range examined, 1-25mgkg(-1). In contrast, the pharmacokinetics of MRL-II were non-linear following both intravenous and oral administration. 3. The acyl glucuronide (AG) conjugate was the only radioactive component detected in bile from dogs dosed with [14C]MRL-I, whereas bile from dogs dosed with [14C]MRL-II contained varying amounts of both the AG and taurine conjugates. The percentages of the acyl glucuronide and taurine conjugates of [14C]MRL-II in dog bile were dose dependent. A higher percentage of radioactivity was associated with the taurine conjugate (about 41%) following intravenous administration at 0.2mgkg(-1) than at 0.9mgkg(-1) (about 14%) or oral administration at 5 mgkg(-1) (about 6%). The decrease in the percentage of radioactivity associated with the taurine conjugate at 0.9 mgkg(-1) was accompanied by a concomitant increase in the amount of the acyl glucuronide. 4. MRL-I, but not MRL-II, was subject to significant enterohepatic recirculation in dogs. Continuous collection of bile resulted in an 11-fold decrease in the terminal half-life of MRL-I in plasma (1.5 versus 16.6 h), and a 2.4-fold increase in its plasma clearance (4.0 versus 1.7 ml min(-1) kg(-1)) after intravenous administration at 1 mg kg(-1). 5. Collectively, the data suggest that the presence and subsequent saturation of the taurine conjugation pathway might have contributed to the non-linear pharmacokinetics of MRL-II in the dog.