A method was developed for the determination of hemoglobin (Hb) adducts formed by the neurotoxic agent acrylamide and its mutagenic epoxide metabolite glycidamide. The method was based on simultaneous measurements of the cysteine adducts formed by these two agents by means of gas chromatography/mass spectrometry in hydrolyzed hemoglobin samples. Rats were injected ip with acrylamide or glycidamide in doses ranging from 0 to 100 mg/kg body wt, and the hemoglobin adduct levels were determined. The hemoglobin binding index of acrylamide to cysteine was found to be 6400 pmol (g Hb)-1/mumol (kg body wt)-1, higher than for any other substance studied so far in the rat, and 1820 pmol (g Hb)-1/mumol (kg body wt)-1 for glycidamide. In rats injected with acrylamide, formation of adducts of the parent compound was approximately linear with dose (0-100 mg/kg), whereas adducts of the epoxide metabolite glycidamide generated a concave curve, presumably reflecting the Michaelis-Menten kinetics of its formation. On the basis of the rate constants for cysteine adduct formation determined in vitro, the first-order rates of elimination of acrylamide and glycidamide from the blood compartment of rats were estimated to be 0.37 and 0.48 hr-1, respectively, using a linear kinetic model. It was further estimated that the percentage of acrylamide converted to glycidamide in the rat decreased from 51% following administration of 5 mg/kg to 13% after a dose of 100 mg/kg. Subchronic treatment of rats with acrylamide (10 mg/kg/day for 10 days or 3.3 mg/kg/day for 30 days) confirmed that the conversion rate of acrylamide to glycidamide, as determined from hemoglobin adduct formation, is higher at low-administered doses. These findings suggest that dose-rate effects may significantly affect risk estimates of this compound and that different low-dose extrapolation procedures should be employed for effects induced by the parent compound acrylamide and those induced by the metabolite glycidamide.