To clarify the cause of the canine individual variability of plasma concentration after oral administration of GTS-21 [(E)-3-(2,4-dimethoxybenzylidene)-3,4,5,6-tetra-hydro-2,3'-bipyridine dihydrochloride], we evaluated the absorption ratio (F(A)), intestinal availability (F(G)), and hepatic availability (F(H)). The bioavailability (F) was evaluated from the ratio of the area under the plasma concentration versus time curves after oral and intravenous administration. Three isoflurane anaesthetised dogs were fitted with an electromagnetic flow probe attached to the portal vein and cannulated through the portal and the femoral veins. After intraduodenal administration of GTS-21, both plasma concentrations were determined simultaneously. F(A) x F(G) was calculated from the portal-systemic concentration difference taking into consideration the blood-plasma partition ratio. F(A) was calculated from the residual drug contents of the small intestine. F(H) was calculated by dividing F by F(A) x F(G). The F values were 0.072, 0.021, and 0.037, indicating an individual variability of ca. threefold. The F(A) values were close to 1, and the F(G) values ranged from 0.449 to 0.461. Accordingly, the F(H) values were estimated at 0.170, 0.047, and 0.083. GTS-21 was completely absorbed but lost by first-pass effects of passage through the gut wall and liver. The first-pass effect of liver is larger than that of the gut wall, and dominates the individual variability in plasma concentration.