The use of flupirtine, an analgesic, has been restricted in European countries because it causes liver injury in rare cases. Flupirtine is primarily metabolized to D-13223, an acetylamino form. In the process of D-13223 formation, it has been hypothesized that a reactive metabolite is formed which may be involved in flupirtine hepatotoxicity. The purpose of this study was to identify the potential reactive metabolite and the responsible enzymes in the human liver to get a clue to the mechanism of hepatotoxicity. Using recombinant enzymes, we found that D-13223 was formed from flupirtine via hydrolysis by carboxylesterase 2 (CES2) and subsequent acetylation by N-acetyltransferase (NAT) 2. A conjugate of N-acetyl-l-cysteine (NAC), a nucleophile, was detected by incubation of flupirtine with CES2, and the conjugate formation in human liver microsomes was inhibited by CES2 inhibitors, indicating that a reactive metabolite, which may be a quinone diimine, was produced in the process of CES2-mediated hydrolysis of flupirtine. The formation of the NAC conjugate in liver S9 samples from NAT2 slow acetylators was significantly higher than that from NAT2 rapid/intermediate acetylators, indicating that NAT2 could function as a detoxification enzyme for flupirtine. CES2-overexpressing HepG2 cells showed remarkable lactate dehydrogenase leakage under flupirtine treatment, while no cytotoxicity was observed in control cells, suggesting that the reactive metabolite formed by CES2-mediated hydrolysis of flupirtine would be a trigger of hepatotoxicity. NAT2 slow acetylators with high CES2 activity could be highly susceptible to flupirtine-induced liver injury.