Myeloperoxidase of phagocytic leucocytes is thought to utilize H2O2 to oxidize halides, which then react with and kill ingested microbes. This hypothesis was based largely on the incorporation of radiolabelled iodide into cells that had phagocytosed bacteria. The present studies investigated the stoichiometry of these reactions and the subcellular localization and electrophoretic pattern of the cellular components that became iodinated. 1. The stoichiometry of the reactions are such that only a small proportion (less than 0.3%) of the total oxygen consumed is utilized for iodination. Iodination after stimulation with the soluble stimulus phorbol myristate acetate (PMA), which is not known to involve the azurophil granules and their contained myeloperoxidase, was comparable with that occurring after bacterial ingestion. 2. Analytical subcellular fractionation of cells that had phagocytosed bacteria localized about 25% of the radioactivity to the membranes, and most of the residual radioactivity distributed with the bacteria and dense granules. In cells stimulated with PMA, more of the radioactivity was associated with the membranes, but about half was still associated with the dense granules. 3. Autoradiographs after dodecyl sulphate/polyacrylamide-gel electrophoresis of cells stimulated with opsonized bacteria gave a similar distribution of iodinated components to that obtained with cells that had been stimulated with PMA or iodinated with Iodogen. These patterns of iodination were very different from those obtained when bacteria alone were iodinated with Iodogen or myeloperoxidase and H2O2. Preparations in which bacteria had been phagocytosed did not show evidence of iodination of bacterial proteins or coating opsonins. Thus positive evidence for the iodination of bacteria has not been produced, and the role of iodination in the microbicidal process of neutrophils remains to be established.