It is unquestionable that the major cause of lung cancer is cigarette smoking. p53 mutations are common in lung cancers from smokers but less common in non-smokers. A large fraction of the p53 mutations in lung cancers are G-->T transversions, a type of mutation that is infrequent in other tumors aside from hepatocellular carcinoma. Previous studies have indicated that there is a good correlation between G-->T transversion hotspots in lung cancers and sites of preferential formation of polycyclic aromatic hydrocarbon (PAH) adducts along the p53 gene. The origin of p53 mutations in lung cancer has been questioned by recent reports suggesting that there are no significant differences in p53 mutation spectra between smokers and non-smokers and between lung cancers and non-lung cancers [S.N. Rodin and A.S. Rodin (2000) Human lung cancer and p53: The interplay between mutagenesis and selection. P:roc. Natl Acad. Sci. USA, 97, 12244-12249]. We have re-assessed these issues by using the latest update of the p53 mutation database of the International Agency for Research on Cancer (14 051 entries) as well as recent data from the primary literature on non-smokers. We come to the conclusion that the p53 mutation spectra are different between smokers and non-smokers and that this difference is highly statistically significant (G-->T transversions are 30 versus 10%; P < 0.0001, chi2 test). A similar difference is seen between lung cancers and non-lung cancers. At a number of mutational hotspots common to all cancers, a large fraction of the mutations are G-->T transversions in lung cancers but are almost exclusively G-->A transitions in non-lung cancers. Our data reinforce the notion that p53 mutations in lung cancers can be attributed to direct DNA damage from cigarette smoke carcinogens rather than to selection of pre-existing endogenous mutations.