Among exocyclic DNA adducts, etheno (epsilon) bases (epsilond A, epsilond C, N(2),3-epsilond G) are generated by reactions of DNA bases with lipid peroxidation (LPO) products derived from endogenous sources and from the carcinogens vinyl chloride or urethane. The recent development of ultrasensitive methods has made it possible to detect these epsilon-adducts in vivo and to study their formation and role in experimental and human carcinogenesis. The promutagenic epsilon-DNA modifications can be detected by immunoaffinity/32P-postlabelling or by immunohistochemistry. When epsilon-adducts are excised from tissue DNA, the modified nucleosides can be quantified in urine by an immunoaffinity-HPLC-fluorescence method. Highly variable background levels of epsilon-adducts were detected in tissues from unexposed humans and rodents, suggesting an endogenous pathway of formation from reaction of trans-4-hydroxy-2-nonenal (via its 2,3-epoxide) with DNA bases. Several known cancer risk factors increased the level of these DNA lesions: Elevated epsilon-adducts were found in hepatic DNA from patients with excess metal storage (haemochromatosis, Wilson's disease), resulting in oxidative stress and high risk of liver cancer. Reactive O/N-intermediates generated during inflammatory processes, for example in patients with inflammatory bowel disease (IBD) and familial adenomatous polyposis (FAP) led to the formation of epsilon-adducts likely through peroxynitrite-mediated LPO and/or increased oxidative arachidonic acid metabolism. A high omega-6-polyunsaturated fatty acid (PUFA) diet increased epsilon-DNA adducts in white blood cells (WBC), particularly in female subjects (about 40-fold), while the level of adducted malondialdehyde in deoxyguanosine of WBC-DNA was only moderately elevated. In conclusion, there is now growing evidence that epsilon-adducts were elevated in cancer-prone patients and in rodents (liver, pancreas, colon, skin), suggesting that promutagenic epsilon-adducts, when formed as a consequence of persistent oxidative stress, can drive cells to malignancy. Therefore, biomonitoring of exocyclic DNA adducts offers useful tools: (i) to evaluate the etiological contributions of dietary fats, oxidative stress, and chronic inflammatory/infectious processes; (ii) to verify the efficacy of chemopreventive agents on endogenous DNA damage and cancer risk; and (iii) to gain mechanistic insights into the role of oxidative stress/LPO-derived lesions in the initiation and progression of human cancer.