The carcinogenicity of petroleum products is mainly due to their content of polycyclic aromatic compounds (PACs). These compounds may be activated metabolically and react with DNA to form DNA adducts, which is a critical event in the initiation of cancer. One of the most common techniques for analyzing DNA adducts is (32)P-postlabeling. The chromatographic method often used has been (32)P-TLC (thin-layer chromatography), but the more recently developed (32)P-HPLC (high-performance liquid chromatography) method has shown advantages. The aim of this study was to test the hypothesis that the (32)P-HPLC method has a better ability of detecting DNA adducts derived from petroleum products than (32)P-TLC. It was found that some DNA adducts migrated from the application point in (32)P-TLC in such a way that it is doubtful if they could be detected and quantified properly. It was also found that, when using (32)P-HPLC, it is possible to use the same protocol for substances with a wide variety of DNA adduct forming potential, whereas (32)P-TLC needs to be optimized regarding time of exposure and/or the amount of DNA applied. Further, a pattern of recognition in (32)P-HPLC enables a selective assessment of DNA adducts derived from complex mixtures whereas (32)P-TLC is very limited when analyzing complex mixtures due to poor resolution. With more knowledge about the properties of the most mutagenic DNA adducts in HPLC, it could be possible to know also which pattern corresponds to a mutagenic or carcinogenic oil. Consequently, (32)P-HPLC is a good alternative when assessing the genotoxicity of petroleum products.