The Fenton reaction was investigated, in a medium approximating to that of the extracellular fluid (ECF), by rapid-mixing stopped flow experiments and HPLC analysis using sodium terephthalate (TA(2-)). The reactive intermediate of the Fenton reaction hydroxylates the essentially nonfluorescent, TA(2-) to the brilliant fluorophor 2-hydroxy-terephthalate (OH-TA), which allows the Fenton reaction to be monitored in stopped-flow experiments. There was no artefactual quenching of the fluorescence by substances present in the Fenton-reaction mixture or in the artificial cerebrospinal fluid (aCSF) that might have influenced OH-TA quantification. A mathematical model based on kinetic considerations was developed. This explains the observed independence of the OH-TA concentration on the amount of TA(2-) present in aCSF as well as its dependence on TA(2-) concentration in potassium acetate buffer. A mechanism based on this model, involving complex formation between Fe(II), TA(2-) and H(2)O(2), followed by an intra-molecular hydroxylation accompanied by an intra-molecular electron transfer was developed. The results are consistent with a reactive intermediate, which causes oxidative stress in vivo, not being a free hydroxyl radical, but a ferryl species or a "crypto" radical. The biological implications of these results are discussed.