A new analytical procedure has been developed for the simultaneous determination of human hemoglobin adducts from aromatic amines and tobacco-specific nitrosamines. These tobacco-related hemoglobin adducts were determined in nonsmokers, smokers, and users of nasal snuff. Adducts from aminobiphenyl compounds are good biomarkers of exposure to tobacco smoke; they are not elevated in users of nasal snuff. However, a significant contribution of environmental exposure to aromatic amines and/or the corresponding nitroaromatics makes it difficult to evaluate passive exposure to tobacco smoke. The best biomarkers for exposure to tobacco smoke should in theory be adducts arising from tobacco-specific nitrosamines. The common adduct from N'-nitrosonornicotine and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone releases 4-hydroxy-1-(3-pyridyl)-1-butanone from hemoglobin upon mild alkaline hydrolysis and only marginal differences are found in the adduct level in smokers and nonsmokers. The reason for this observation is not yet understood and is currently under investigation. However, the adduct formed by tobacco-specific nitrosamines is well suited for the detection of oral and nasal tobacco use. Only by simultaneous determination of both adducts formed by aromatic amines and tobacco-specific nitrosamines is it possible to differentiate between nonsmokers, smokers, and nasal snuff users.