In recent years it has been apparent that many of the known antiparasitic drugs produce free radicals. Intracellular reduction followed by autooxidation yielding O.-2 and H2O2 has been suggested as the mode of action of nifurtimox on Trypanosoma cruzi and as the basis of its toxicity in mammals. On the other hand, free radical intermediates that do not generate oxygen-reduction products under physiological conditions have been found in the metabolic pathways of other antiparasitic nitro compounds (benznidazole, metronidazole, and other 5-nitroimidazoles) used in the treatment of diseases such as Chagas' disease, trichomoniasis, giardiasis, balantidiasis, amebiasis, and schistosomiasis. In these cases, as well as in the case of niridazole (used in the treatment of schistosomiasis), covalent binding or other interactions of the intermediates of nitroreduction with parasite macromolecules are possibly involved in their toxicity. Redox cycling of these compounds under aerobic conditions appears to be a detoxification reaction by inhibiting net reduction of the drugs.