Biotransformation of styrene and its toxic metabolite, phenyloxirane (1), in mice in vivo was studied. Mice were treated with single intraperitoneal doses of styrene (400 mg/kg of body weight), and with (R)-, (S)-, or racemic styrene oxide (150 mg/kg of body weight). Profiles of neutral and acidic metabolites were determined by GC/MS. Mandelic acid (3) and two mercapturic acids, N-acetyl-S-(2-hydroxy-2-phenylethyl)cysteine (5) and N-acetyl-S-(2-hydroxy-1-phenylethyl)cysteine (6), were found to be major urinary metabolites of both styrene and phenyloxirane. 1-Phenylethane-1,2-diol (2) was the main neutral metabolite. The rate of excretion of this metabolite, as determined by GC, was 5-10 times lower than that of mandelic acid. Several minor acidic metabolites were also identified. Among them, novel phenolic metabolites, namely, 2-(4-hydroxyphenyl)ethanol (7), (4-hydroxyphenyl)acetic acid (11), and two isomeric hydroxymandelic acids (12), are of toxicological significance. Main stereogenic metabolites were isolated as methyl esters from extracts of pooled acidified urine treated with diazomethane. The mandelic acid that was obtained was converted to diastereomeric Mosher's derivatives prior to analysis by NMR. Mercapturic acids were analyzed directly by (13)C NMR. Pure enantiomers of 1 were metabolized predominantly but not exclusively to corresponding enantiomers of 3. Styrene yielded predominantly (S)-mandelic acid. Fractions of mercapturic acids 5 and 6 isolated from urine amounted to 12-15% of the dose for all compounds that were administered. Conversion to mercapturic acids was highly regio- and stereoselective, yielding predominantly regioisomer 5. Styrene, as compared to racemic phenyloxirane, yielded slightly more diastereomers arising from (S)-1 than from (R)-1. These data can be explained by formation of a moderate excess of the less mutagenic (S)-1 in the metabolic activation of styrene in mice in vivo.