The oxidation of 2',7'-dichlorofluorescin (DCFH) and its diacetate form (DCFHDA) by the HRP/peroxynitrite system was investigated. Both DCFH and DCFHDA were oxidized to fluorescent products. A major anomaly, however, was the observation that fluorescence continued to build up long after peroxynitrite total decomposition and the initial HRP compound I reduction, suggesting the production of oxidants by the system. Indeed, preformed HRP compound I was instantly reduced by DCFH and DCFHDA to compound II with the obligate formation of DCF(-) semiquinone and DCFHDA-derived radicals. Catalase strongly inhibited fluorescence and EPR signals, suggesting the intermediate formation of H2O2. Taken together the data indicate that peroxynitrite rapidly oxidizes HRP to HRP compound I, which is reduced by DCFH and its diacetate form with the concomitant formation of DCF(-) semiquinone and DCFHDA-derived radicals. These are oxidized by O2, producing O2(-) (as demonstrated by EPR and oxygen consumption experiments), which dismutates to produce H2O2, which serves to fuel further DCFH/DCFHDA oxidation via HRP catalysis. Also DCFHDA was shown to be considerably more resistant to oxidation than its hydrolyzed product DCFH, presumably because of the absence of the easily oxidizable phenol moieties. DCFHDA/DCFH have been used to study free radical production in a variety of systems. Our findings demonstrate that this assay is subject to a serious artifact in that it produces what it is purported to measure; therefore, its use in biological systems should be approached with caution.