The reaction of ferrous iron with hydrogen peroxide (the Fenton reaction) has been widely used as a means of inducing strand cleavage in DNA, generally for the purpose of "footprinting" protein-DNA complexes or studying DNA structure. The identity of the DNA-oxidizing species produced in the reaction, however, has been an issue of debate. By comparing gamma-radiolysis-induced DNA cleavage patterns in a variety of buffers with those generated by the reaction of [Fe(EDTA)](2-) + H2O2 in the presence of ascorbate, we report that under the conditions used for a typical footprinting experiment, results are consistent with production of the hydroxyl radical as the oxidant responsible for DNA strand scission. We also cite experiments that argue against the participation of a high-valent iron-ore complex in DNA cutting.