1. The study investigated mechanisms underlying the pharmacokinetic differences of two zwitterionic diastereomers ((3S)-3-[(3R or 3S)-2-oxo-3-[3-(5,6,7,8-tetrahydro-1,8-naphthyridin-2-yl)propyl]pyrrolidin-1-yl]-3-quinolin-3-ylpropanoic acid) with different lipophilicities using a combination of in vivo and in vitro approaches. 2. In rat, both isomers possessed comparable plasma clearances (CL). However, the more lipophilic diastereomer I exhibited a higher metabolic clearance (>2-fold higher than II), whereas the hydrophilic zwitterion II exhibited a higher biliary clearance (approximately 5-fold higher than I). Following oral administration, the bioavailability (F) of I (17%) was much higher than that of II (1%). 3. Consistent with these in vivo observations and the expectation based on their lipophilicity differences, the metabolism in rat liver microsomes was faster and the permeability in Caco-2 and LLC-PK1 cells and in situ rat intestinal loop was better for I than for II. 4. Only the absorption of the more lipophilic diastereomer I was subjected to an efflux system in the Caco-2 and in situ rat intestinal loop models. I was a good substrate for P-glycoprotein (P-gp) in both the human MDR1 and mouse mdr1a transfected cell lines, and in the wild-type mdr1a (-/-) mouse when compared with the P-gp-deficient mdr1a (-/-) mouse. Concomitant administration of I with verapamil in rat caused significant increases in oral AUC, F and Cmax of I without affecting its CL, further supporting the effect of P-gp in limiting the intestinal absorption of I in vivo in this animal model. 5. Since the findings that the lipophilic diastereomer I, but not II, was a good P-gp substrate were not in line with the observations that I was excreted to bile much slower than II and that I was absorbed better than II, the results suggested that P-gp played a minor role to the observed differences in the biliary excretion and intestinal absorption of the diastereomers I and II in rat.