In a previous study of the metabolism of methyl-n-amylnitrosamine (MNAN) in the rat, 2- to 5-hydroxy-MNAN (HO-MNAN) were provisionally identified as metabolites and the identity of 4-HO-MNAN was confirmed by mass spectrometry. We now describe syntheses and mass and other spectra for 2- to 5-oxo-MNAN. Two previously unidentified MNAN metabolites were shown to be 3- and 4-oxo-MNAN. In addition to 4-HO-MNAN, we confirmed 3-HO-, 4-oxo- and (less certainly) 2-HO-MNAN as urinary MNAN metabolites by GLC-MS of HPLC fractions. Analysis with and without beta-glucuronidase treatment showed that the urinary HO-MNANs occurred as their beta-glucuronides. MNAN (25 mg/kg injected i.p.) had a blood half-life of 21 min in adult male rats. The blood also contained 4-HO- and 4-oxo-MNAN, which showed maximum levels that were 13 and 26% respectively of that for MNAN, and were cleared more slowly than MNAN. On incubation for 3 h with MNAN, rat esophagus produced 3- and 4-oxo-MNAN in yields that were 5% of those for the corresponding HO-MNANs. For MNAN metabolism, the 4-oxo-/4-HO-MNAN ratio of metabolites was 5% for adult rat liver and was 22% for adult hamster liver and 9-day-old rat liver. On incubation with 4-HO-MNAN for 3 h, oxidation to 4-oxo-MNAN was 16-25% for adult hamster or 9-day-old rat liver slices and for adult hamster liver homogenate. Homogenate activity was concentrated in the microsomal fraction, for which NAD was a more effective co-factor than NADP. A bacterial alcohol dehydrogenase oxidized 4-HO- to 4-oxo-MNAN in 38% yield/3 h. None of these preparations oxidized 2-HO- to 2-oxo-MNAN. It was concluded that 3- and 4-oxo-MNAN were metabolites of MNAN, apparently (for 4-oxo-MNAN) via HO-MNAN oxidation by a microsomal NAD-dependent enzyme, that 4-HO- and 4-oxo-MNAN formation was a major route of MNAN metabolism, and that 4-oxo-MNAN might play a role in MNAN carcinogenesis.