Human lung tissue is frequently studied as a target organ for DNA damage from carcinogen-DNA adducts. In order to assess the distribution of carcinogen-DNA adducts in human lung, we measured 7-methyl-2'-deoxyguanosine-3'-monophosphate (7-methyl-dGp), 7-ethyl-2'-deoxyguanosine-3'-monophosphate (7-ethyl-dGp) and 4-hydroxy-(3-pyridyl)-1-butanone (HPB)-releasing DNA adducts in different lobes. The first two result from exposure to N-nitrosamines, including tobacco-specific nitrosamines, and the latter only from tobacco-specific nitrosamines. Using a chemically-specific 32P-postlabeling assay for 7-alkyl-2'-deoxyguanosines, adducts were measured in eight separate lung segments of ten autopsy donors. 7-Methyl-dGp levels were detected in all eighty samples (range from 0.3 to 11.5 adducts/10(7) dGp; mean 2.5 +/- 2.3 adducts/10(7) dGp). 7-Ethyl-dGp were detected in all but five of the samples (range from <0.1 to 7.1 adducts/10(7) dGp; mean 1.6 +/- 1.7 adducts/10(7) dGp). 7-Methyl-dGp levels were approximately 1.5-fold higher than 7-ethyl-dGp levels, and they were positively correlated with each other in most individuals. There was no consistent pattern of adduct distribution in the different lobar segments. Most individuals, especially those with the lowest levels, had similar levels among the lobes, while those with the highest levels had a widely variable pattern ranging as much as ten-fold. Moreover 7-methyl-dGp and 7-ethyl-dGp levels in all people showed a highly significant inter-individual variation (P = 0.0001). The levels of 7-alkyl-2'-deoxyguanosine among individuals could not be explained by differences in tobacco exposure (measured by serum cotinine), onset of death, gender, age, race, blood ethanol, or ventilation and perfusion variability. In an effort to corroborate 7-alkyl-2'-deoxyguanosine adducts variability among lobes or individuals, we sought to determine a correlation with HPB-releasing DNA adducts as an independent marker of tobacco exposure. However, this tobacco- specific carcinogen-DNA adduct could not be detected in four individuals tested (detection limit: 0.3 adducts per 10(7) dGp). Based upon the lack of 7-alkyl-2'-deoxyguanosine discernible adduct patterns, no conclusions could be drawn regarding a potential relationship to lobar cancer incidence. The results indicate that in studies of carcinogen-DNA adducts, such as 7-alkyl-dGp in human lungs, for most individuals a random lung sample would be representative of other parts of the lungs. Some individuals however might be misclassified due to highly variable 7-alkyl-dGp levels.