Serum lipid peroxidation was measured by three different methods (diene conjugation, DC; thiobarbituric acid reactive material, TBARM; and fluorescent chromolipids, FC) during physical exercise of different duration. The aim of the study was to investigate which of the three methods is the most representative in measuring oxidative stress after physiological stimulus (physical exercise) in man. In general, diene conjugation is a measure of early events of lipid peroxidation reactions, TBARM measures end products of lipid peroxidation and FC are products formed by reaction of end products of lipid peroxidation with the amino groups. In trial I, 8 subjects ran 1 km with maximal speed; in trial II, 7 subjects ran 10 km with maximal speed; in trial III, 9 subjects ran 27 km with maximal speed, and in trial IV, 5 subjects cycled 8 days in competition, altogether 1061 km biking. In each trial venous blood samples were taken before and immediately after the exercise. In trials II, III, and IV serum DC concentration increased by 14% (P < 0.05), 11% (P < 0.01), and 25% (P < 0.05), respectively, during the exercise. FC or TBARM did not change in the trials. We conclude that DC as a measure of the early phase of lipid peroxidation may be less affected by protective antioxidant functions than FC and TBARM methods, which both measure end products of lipid peroxidation. Therefore, DC may be the most sensitive of the three methods to estimate serum lipid peroxidation induced by exercise in man. We also suggest that the apparently conflicting results of serum lipid peroxidation in earlier studies may be caused by use of different methods to measure lipid peroxidation.