Epidemiological investigations repeatedly show decreased morbidity from regular exercise compared with sedentary life. A large number of investigations have demonstrated increased oxidation of important cellular macromolecules, whereas other investigators have found no effects or even signs of lowering of oxidation of macromolecules. In particular, extreme and long-duration strenuous exercise appears to lead to deleterious oxidation of cellular macromolecules. The oxidation of DNA is important because the oxidative modifications of DNA bases, particularly the 8-hydroxylation of guanine, are mutagenic and have been implicated in a variety of diseases such as ageing and cancer. The methodologies for further investigation of the relationship between DNA oxidation and exercise are available. The preferred methods rely on HPLC or GC-mass spectrometry; whereas the theoretically-attractive liquid chromatography- tandem mass spectrometry is being developed. Caution should be taken to avoid artifacts because of the six orders of magnitude of difference between oxidized and non-oxidized DNA bases in tissues. The methods can be used to estimate tissue levels, i.e. a local concentration of oxidized DNA, or to estimate the rate of body DNA oxidation by the urinary output of repair products, the latter being a method that is independent of repair. During exercise there appears to be a shifting of dietary-dependent antioxidant, e.g. vitamin C and vitamin E, from muscle to plasma, and an increased oxidation in plasma of these antioxidants. Supplementation trials with antioxidants have not been able to increase exercise performance; however, optimum nutrition with antioxidants and possibly supplementation, could be important in the prevention of diseases in the long term. The pattern from these observations appears to be quite consistent; immediately after exercise, regardless of how intense, there do not appear to be any signs of oxidative damage to DNA. Acute or prolonged moderate exercise does not produce signs of oxidative DNA damage and might even be associated with lowering of the levels of oxidation of tissue DNA; however, after long-duration and intense exercise an increase in oxidative DNA modifications is apparent. We suggest as a hypothesis that the relationship between exercise and health is U-shaped. This hypothesis needs to be tested in detail in order to establish the maximum beneficial exercise level with regard to oxidative DNA modification, and also the level that could be deleterious and might even increase the risk for cancer and other diseases.