This article provides a comprehensive analysis of the redox reaction between glutathione/glutathione disulfide and ascorbic acid/dehydroascorbic acid. It includes an historical perspective of the progression of the experiments, first begun more than 60 years ago and continuing today with heightened importance. Indeed, the antioxidant capacity of glutathione and ascorbic acid, whether singly or in combination, linked via the redox couple, is a subject of intense interest for studies by bench scientists and clinicians, particularly because a growing body of evidence suggests that free radicals may be involved in a variety of diseases. The authors begin with a detailed summary of "test tube" experiments (the chemical perspective) that have revealed the conditions that regulate the rate of the redox coupling between glutathione and dehydroascorbic acid and that promote or inhibit the decomposition of dehydroascorbic acid in ordinary, buffered aqueous media; results obtained in the authors' laboratory are used for illustration purposes and uniformity of presentation. The authors then proceed to a critical examination of the extent to which the redox couple between glutathione and ascorbic acid operates in a cell, using the often published antioxidant cascade (See Fig. 1) as the model for the analysis (the physiological perspective). The evidence for and the evidence against the presence of the enzyme dehydroascorbate reductase in animal cells is outlined in a balanced way in an attempt to make sense of this continuing controversy. Next, the authors carefully document the many studies showing that exogenous dehydroascorbic acid is transported into cells where it is reduced to ascorbic acid by glutathione. Finally, they probe the functional significance and efficiency of the redox couple in monolayer cultures of human retinal pigment epithelial (RPE) cells, as a prototypical cellular model. The authors include the results of new experiments showing that incubation of RPE cells with a nitroxide, TEMPOL, leads to the selective oxidation of intracellular ascorbic acid. This approach is desirable because it dissects the cascade at a specific site and permits measurements of the levels of ascorbic acid and glutathione in the cells before, during, and after oxidation. The results show that only partial regeneration of ascorbic acid is obtained when control conditions are restored. However, if either ascorbic acid or dehydroascorbic acid is added to the media during the recovery period following treatment of cells with TEMPOL, then full recovery of ascorbic acid is observed.(ABSTRACT TRUNCATED AT 400 WORDS)