concerning the study of Wang et al. 2 taught the readers of HEPATOLOGY a great deal about the mechanism of acetaminophen (AP)-induced hepatotoxicity. However, as the corresponding author for the paper of Wang et al., I feel obliged to respond to his comments concerning the dose of acarbose and ethanol used in our study.Dr. Krähenbü hl stated that the daily dose of alcohol and acarbose ingested by our rats, approximately 14 mL of 10% alcohol and 16 mg acarbose per day, corresponded to approximately 200 g of alcohol and 3 g of acarbose by a 70 kg human subject, and concluded that the doses of both ethanol and acarbose that we chose were substantially higher than those normally ingested. 1 However, the extrapolation of the dose from rats to humans can also be made on the basis of body surface area, rather than on the conventional body weight basis. 3 In Wang et al., 2 21 to 22 mg/kg/d of acarbose evidently potentiated the hepatotoxicity of CC1 4 and/or AP, which was significantly increased when combined with 3.0 g/kg/d alcohol. This dose of acarbose is about four times higher than the recommended dose (5 mg/kg/d) for type 2 diabetics when calculated on the basis of body weight, which is a substantially high dose, as Dr. Krähenbü hl remarks. 1 However, the dose extrapolation based on body surface area produces a different result. Assuming that the body surface area is proportional to body weight (kg) to the power of 0.7, 4 i.e., (body weight) 0.7 , the doses of acarbose (21.5 mg/kg) and alcohol (3.0 g/kg) for a rat correspond to doses of only 421 mg of acarbose and 58.7 g of alcohol for a 70-kg subject. Therefore, by this measure, the dose of acarbose (21.5 mg/kg) used by Wang et al. is roughly the same as that normally taken by diabetics. Furthermore, the dose of alcohol (3.0 g/kg) that the rats received corresponds to a moderate human intake, an amount that is ingested by a considerable part of the population. No firm conclusion has been made as to whether the dose extrapolation is more appropriately based on body surface area or on body weight when extrapolating a finding from rats to man. We need, therefore, only to focus on the finding that acarbose alone or in combination with alcohol potentiates AP hepatotoxicity by affecting the bioactivation of AP in the liver. Patients who regularly take ␣-glucosidase inhibitors such as acarbose should be warned to avoid overusing drugs that exert their toxic effects through a bioactivation mechanism such as AP.