A liquid chromatography-atmospheric pressure chemical ionization-tandem mass spectrometry (LC-APCI-MS/MS) method was developed to analyze human urine for 4-oxo-4-(3-pyridyl)butanoic acid (keto acid) and the enantiomers of 4-hydroxy-4-(3-pyridyl)butanoic acid (hydroxy acid) to test our hypothesis that (S)-hydroxy acid could be a biomarker of metabolic activation of the tobacco-specific carcinogens 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and N'-nitrosonornicotine (NNN) while (R)-hydroxy acid would be formed predominantly from nicotine, as indicated by studies with rats. Urine was collected from smokers, and from the same individuals after they had stopped smoking and used a nicotine transdermal system (nicotine patch) for 3 weeks. If (S)-hydroxy acid were a biomarker of NNK and NNN metabolic activation, its levels should be higher in the urine of smokers than in nicotine patch users because tobacco smoke, but not the nicotine patch, contains NNK and NNN. Internal standard, [2,2,3,3,4-D5]hydroxy acid, was added to an aliquot of urine, which was then subjected to solid phase extraction. The eluant containing hydroxy acid was esterified with acidic methanol, followed by treatment with (S)-(-)-alpha-methylbenzyl isocyanate, producing methyl-4(S)- or methyl-4(R)-[(S)-alpha-methylbenzylcarbamoyl]-4-(3-pyridyl)buta noate [(S,S)- or (R,S)-MMPB, respectively]. After HPLC purification, the MMPB diastereomers were separated and quantified by LC-APCI-MS/MS. Mean levels of (S)- and (R)-hydroxy acid were 14.1 +/- 8.0 and 1120 +/- 600 ng/mL, respectively, in smokers during ad lib smoking (n = 18), while the corresponding levels during nicotine patch use (n = 18) were 4.1 +/- 3.3 and 363 +/- 228 ng/mL. The amounts of (S)-hydroxy acid were far higher than could be formed from NNK and NNN, and the total amount of hydroxy acid indicated that it was a substantial urinary metabolite of nicotine, in contrast to results with rats. Therefore, the study was extended to quantify keto acid. This was accomplished by NaBH4 treatment of urine, which converted keto acid to hydroxy acid quantitatively, which was in turn analyzed as described above. Levels of keto acid while subjects were smoking and using the nicotine patch were 228 +/- 129 (n = 8) and 97.5 +/- 80.6 ng/mL (n = 8), respectively. These results indicate that conversion of nicotine to keto acid and hydroxy acid is a substantial metabolic pathway in humans, accounting for an estimated 14% of the nicotine dose. Apparently, keto acid is extensively converted to hydroxy acid in humans, in contrast to the results with rats. (S)-Hydroxy acid in human urine cannot be used as a biomarker of NNK and NNN metabolic activation because it is overwhelmed by the (S)-hydroxy acid formed from nicotine, despite the fact that >98% of the urinary hydroxy acid has the (R)-configuration. These results provide new insights about nicotine metabolism in humans.