Normal and cytochalasin B-treated human granulocytes have been studied to determine some of the interrelationships between phagocytosis-induced respiration and superoxide and hydrogen peroxide formation and release into the extracellular medium by intact cells. By using the scopoletin fluorescent assay to continuously monitor extracellular hydrogen peroxide concentrations during contact of cells with opsonized staphylococci, it was demonstrated that the superoxide scavengers ferricytochrome c and nitroblue tetrazolium significantly reduced the amount of H(2)O(2) released with time from normal cells but did not abolish it. This inhibitory effect was reversed by the simultaneous addition of superoxide dismutase (SOD), whereas the addition of SOD alone increased the amount of detectable H(2)O(2) in the medium. The addition of sodium azide markedly inhibited myeloperoxidase-H(2)O(2)-dependent protein iodination and more than doubled H(2)O(2) release, including the residual amount remaining after exposure of the cells to ferricytochrome c, suggesting its origin from an intracellular pool shared by several pathways for H(2)O(2) catabolism. When cells were pretreated with cytochalasin B and opsonized bacteria added, reduced oxygen consumption was observed, but this was in parallel to a reduction in specific binding of organisms to the cells when compared to normal. Under the influence of inhibited phagosome formation by cytochalasin B, the cells released an increased amount of superoxide and peroxide into the extracellular medium relative to oxygen consumption, and all detectable peroxide release could be inhibited by the addition of ferricytochrome c. Decreased H(2)O(2) production in the presence of this compound could not be ascribed to diminished bacterial binding, decreased oxidase activity, or increased H(2)O(2) catabolism and was reversed by the simultaneous addition of SOD. Furthermore, SOD and ferricytochrome c had similar effects on another H(2)O(2)-dependent reaction, protein iodination, in both normal and cytochalasin B cells. When oxygen consumption, O(2.) (-), and H(2)O(2) release were compared in the presence of azide under identical incubation conditions, the molar relationships for normal cells were 1.00:0.34:0.51 and for cytochalasin B-treated cells 1.00:0.99:0.40, respectively. Nonopsonized, or opsonized but disrupted, bacteria did not stimulate any of these metabolic functions. The results indicate that with normal cells approximately 50% of H(2)O(2) released during phagocytosis is derived directly from O(2.) (-) by dismutation, the remainder appearing from an (intra)cellular source shared by azide-inhibitable heme enzymes. With cytochalasin B treatment the evidence is consistent with the derivation of all H(2)O(2) from an O(2.) (-) precursor which is released from the cell surface. Furthermore, when activated by phagocytic particle binding, the neutrophil O(2.) (-) generating system appears to make more of this compound than can be accounted for by dismutation to H(2)O(2). This establishes conditions for the direct participation of both compounds in the microbicidal and cytocidal activity of these cells.