The mechanism of formation of etheno (epsilon) adducts of nucleic acid bases from 2-haloacetaldehydes is generally assumed to occur via initial Schiff base formation resulting from reaction of the aldehyde with an exocyclic amine. We recently revised the 1H NMR assignments of the epsilon protons of 1,N2-epsilon-Guo (Guengerich, F. P., Persmark, M. P., and Humphreys, W. G. (1993) Chem. Res. Toxicol. 6, 635-648). In that work we also observed a facile and specific exchange of H7 of 1,N2-epsilon-Guo and H5 of N2,3-epsilon-Gua with H2O. These findings raise questions about the mechanistic conclusions reached on the basis of labeling studies with deuterated ClCH2CHO (Sattsangi, P. D., Leonard, N. J., and Frihart, C. R. (1977) J. Org. Chem. 42, 3292-3296). BrCH2-13CHO was prepared from BrCH2(13)CO2H and used to prepare 1,N2-epsilon-Guo (from Guo) and O6-ethyl-N2,3-epsilon-Gua (from O6-ethylGua). The positions of the labels were determined by 1H NMR spectroscopy experiments to be adjacent to the original Gua N2 (exocyclic) atom in both cases, i.e., at C6 in both epsilon products. The labeling patterns are consistent with a mechanism involving initial Schiff base formation from the N2 atom and the aldehyde and subsequent nucleophilic attack of an endocyclic nitrogen on the methylene carbon.