We undertook this study to answer several questions regarding nitrosamine metabolism. Kinetics of nitrosamine metabolism showed the involvement of at least two enzymes in the dealkylation of N-nitrosodiethylamine (NDEA) and N-nitrosodimethylamine (NDMA) in mouse liver microsomes. Coumarin inhibited both reactions competitively. On the other hand, microsomal coumarin 7-hydroxylase was inhibited by NDMA (Ki 2.7 mM) and NDEA (Ki 0.013 mM). The big difference in the Ki values suggests a higher affinity of NDEA than NDMA to Cyp2a-5 (mouse cytochrome P450coh). A specific antibody against Cyp2a-5 inhibited more of the microsomal NDEA (up to 90%) than NDMA (up to 40%) dealkylation. The converse was true with anti-Cyp2e-1 antibody. These results suggest that the primary substrate for Cyp2a-5 is NDEA and for Cyp2e-1, NDMA. Western blot analysis of human liver microsomes showed a great interindividual variation in the amounts of CYP2A6 (human cytochrome P450coh) and CYP2E1. Also, coumarin 7-hydroxylation and nitrosamine dealkylation varied greatly among individuals. A high correlation (r = 0.93, P < 0.001) was found between NDEA and coumarin metabolism. Both activities were associated with CYP2A6. On the other hand, little or no correlation was found between microsomal CYP2A6 and CYP2E1 or between CYP2E1 and NDEA dealkylation. Immunoinhibition of human microsomal NDEA metabolism by CYP2a-5 antibody varied greatly among individuals (10-90%), suggesting, as in the case of mice, that NDEA is metabolized primarily by CYP2A6, at least in some individuals. Taken together the data suggest that (1) the metabolic activation of nitrosamines in humans varies greatly among individuals; (2) different nitrosamines may partially be metabolized by different cytochrome P450 isozymes; and (3) because of similarities between nitrosamine metabolism in mice and humans, inbred strains of mice would be relevant experimental models for studying nitrosamine activation.