To clarify the cause of the canine individual variability in plasma concentration after oral administration of GTS-21, we evaluated in vitro the metabolism to 4-OH-GTS-21 in liver microsomes of the same individuals from in vivo pharmacokinetic study. First, we applied to the Michaelis-Menten kinetic parameters to a dispersion model, and compared hepatic availability (F(H)) and hepatic clearance (CL(H)) values from in vitro with bioavailability (F), hepatic plasma flow (Q(PH)), and plasma clearance (CL(P)) values from in vivo. The ratios of CL(H) to Q(PH) were ranged 0.74 to 0.94, suggesting that GTS-21 is a hepatic plasma flow-limiting drug. A significant correlation of F(H) and F in the four dogs (r=0.995, p=0.005) indicates that the variability is predominantly caused by GTS-21 O4-demethylase activity. Second, we specified the cytochrome P450 (CYP) enzymes that are involved with the metabolism by chemical inhibition. alpha-Naphthoflavone, furafylline, quinidine, quinine, and troleandomycin significantly inhibited GTS-21 O4-demethylase activity. Thus CYP1A, CYP2D15, and CYP3A12 were involved with O4-demethylation. The variability in control activity decreased on addition of alpha-naphthoflavone and furafylline. Third, we quantified the contents of CYP1A and CYP3A12 by enzyme-linked immunosorbent assay. The content of CYP1A was consistent with GTS-21 O4-demethylase activity. We concluded that canine liver CYP1A causes the individual variability in GTS-21 plasma concentration after oral administration.