Exposure to the tobacco-specific N-nitrosamine 4-(methylnitrosamino)-1-(3- pyridyl)-1-butanone (NNK) is considered to be an important etiological risk factor for lung cancer in tobacco users. The metabolism of NNK via carbonyl reduction to 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), α-hydroxylation to form both DNA methylating and pyridyloxobutylating intermediates, and detoxification by pyridyl N-oxidation and glucuronide formation are well-characterized in laboratory animals but less so in man. The in vitro kinetics of 0.03-250 μM [5-3H]NNK metabolism were determined under identical experimental conditions using female A/J mouse, male Fischer 344 rat, female Syrian golden hamster, and human lung tissue explants in tissue culture. The concentration-dependent percentage contribution of the three major pathways of NNK metabolism (carbonyl reduction, R-hydroxylation, and N-oxidation) showed large interspecies variation. Quantitatively, in mouse, carbonyl reduction to NNAL increased steadily with an increasing substrate concentration (10-74% total NNK metabolism), while concurrent decreases occurred in end products of α-hydroxylation (60 to 18%) and N-oxidation (42 to 5%). In rat lung, there were no apparent concentration-dependent trends (NNAL, 42 ± 4%; α-hydroxylation, 35 ± 2%; and N-oxidation, 24 ± 3%). In hamster lung, a clear concentration-dependent increase in the contribution of NNAL to total NNK metabolism (from 47 to 87%) was paralleled by a steady decline in end products of α-hydroxylation (31 to 11%) and N-oxidation (22 to 2%). Human lung metabolism showed no concentration-dependent tendencies (NNAL, 89 ± 1%; α-hydroxylation, 8.8 ± 1.1%; and N-oxidation, 2.1 ± 0.3%). The major α-hydroxylation product in human lung was 4-hydroxy-1-(3-pyridyl)-1-butanone (keto alcohol), thus supporting the potential pyridyloxobutylation of lung DNA. Metabolism to 4-(3-pyridyl)-4-oxobutanoic acid (keto acid), which could result in lung DNA methylation, was only sporadically seen in human lung but present to a far greater extent in rodent lung. No evidence for glucuronidation was found in any species. Generally, the rate of formation of all NNK metabolites showed two different enzyme kinetics, resulting in large differences between apparent Km and Vmax values in the low (up to 2.8 μM) and high substrate concentration ranges. The metabolism of NNK by α-hydroxylation is considerably lower in human lung as compared to that observed in rodent species, suggesting that extrapolation of in vitro rodent data to man may result in invalid conclusions about the capacity of the human lung to activate NNK under realistic conditions of NNK exposure expected to occur in man.