When the esophageal carcinogen methyl-n-amylnitrosamine (MNAN; concentration, 3 mg/liter) was incubated in vitro with rat esophagi for 3 hr, five principal neutral metabolites (Metabolites 2 to 6; total yield, 3.0% of the MNAN per 100 mg tissue) were separated by gas chromatography, with detection by a thermal energy analyzer. Rat liver produced similar metabolites (total yield, 2.1% of the MNAN per 100 mg tissue). Metabolites 4 to 6 and a minor product, Metabolite 7, were tentatively identified as 2-, 3-, 4-, and 5-hydroxy-MNAN (HO-MNAN), respectively, from a comparison of their gas chromatography retention times with those of the synthesized compounds. Rat esophagus produced similar amounts of 3- and 4-HO-MNAN and lesser amounts of 2-HO-MNAN, whereas rat liver produced mainly 4-HO-MNAN. Rat nasal tissue metabolized 8.0% of the MNAN per 100 mg tissue, with a metabolite pattern like that of the esophagus. Rat lungs produced mostly 5-HO-MNAN. A comparison of yields from tissues of rats, hamsters, guinea pigs, and mice supported the view that the total production of neutral MNAN metabolites indicated the sensitivity of MNAN carcinogenesis, with some exceptions. MNAN injected i.p. was less carcinogenic for the esophagus and nasal cavity in Sprague-Dawley than in MRC-Wistar rats, perhaps because the livers of Sprague-Dawley rats metabolized more of the MNAN. The urine of MNAN-treated MRC-Wistar rats contained MNAN and metabolites provisionally identified as 2-, 3-, and (as the major product) 4-HO-MNAN. The identity of the urinary 4-HO-MNAN was confirmed by gas chromatography-mass spectrometry. We speculate that tissues like the esophagus, which (unlike the liver) produce significant proportions of 2- and 3-HO-MNAN, also produce significant amounts of the most likely proximal carcinogen, 1-HO-MNAN.