An in vivo murine model was developed to measure maternal phenytoin biotransformation along with the covalent binding of phenytoin to fetal tissues in the same fetuses which were assessed for fetal anomalies. Acetaminophen was administered to pregnant CD-1 mice 1 hour prior to phenytoin, both given i.p. at varying doses and gestational times between days 11 and 13. Dams were killed between days 12 and 19. Metabolites reflecting the enzymatic bioactivation of phenytoin were quantified in maternal plasma and urine with high-performance liquid chromatography (HPLC). Acetaminophen pretreatment caused a threefold increase in phenytoin-induced fetal cleft palates without increasing resorptions. The covalent binding of radiolabeled phenytoin to fetal and placental tissues measured on day 13 was increased twofold and threefold, respectively, by acetaminophen pretreatment. Phenytoin covalent binding measured on day 16 was significantly increased in the livers of fetuses with cleft palates, but not in the livers of dams with fetuses having cleft palates. Binding to fetal brain on day 16 was over fourfold higher than that in maternal brain. Acetaminophen pretreatment differentiated dams into poor and extensive metabolisers of phenytoin, with only the latter group carrying fetuses with cleft palates. The incidence of fetal cleft palates correlated positively with maternal urinary levels of phenytoin (r = +.81, P less than .01) and its dihydrodiol metabolite (r = +.61, 0.05 less than P less than .1), and negatively with levels of para-hydroxylated phenytoin (r = -.85, P less than .01). These findings related both to the mechanism of phenytoin teratogenicity and its potentiation by acetaminophen.