Lymphoma cells were rapidly lysed by activated macrophages and granulocytes in the presence of PMA. Release of 51Cr from lymphoma cells correlated closely with their destruction as viewed by scanning electron microscopy, and with reduction in the number of trypan blue- excluding cells. The standard assay involved 51 Cr release measured at 4.5 h, but injury appeared to be complete in 1 h. Of eight different types of effector cells tested, only those releasing abundant H2O2 in response to PMA were effective, that, is BCG-, C. parvum-, or casein- activated macrophages, or thioglycollate-elicited granulocytes. Normal macrophages, J774 cells, or macrophages elicited with thioglycollate broth or proteose-peptone were ineffective. BCG-activated macrophages and granulocytes caused 50% specific release of 51Cr from P388 lymphoma cells at E:T ratios between 1.4 and 4.5, and from mouse erythrocytes at E:T ratios of 0.017 to 0.025. 10 types of target cells varied widely in their susceptibility to lysis by reagent H2O2, with one-half maximal lysis occurring at H2O2 concentrations ranging from 3.63 X 10(-6) M to 3.85 X 10(-5) M. Effector cells were expected to generate approximately that much H2O2 during the period of injury. Susceptibility of the target cells to lysis by PMA-triggered granulocytes correlated closely with their sensitivity to H2O2 (r = 0.98). The membrane-active agents LPS and digitonin, which did not trigger H2O2 release, did not trigger cytotoxicity. The dose-response curve for triggering of H2O2 release by PMA was identical to that for triggering cytotoxicity. These results provided strong circumstantial evidence for the importance of H2O2 in extracellular cytolysis by activated macrophages and granulocytes when pharmacologically triggered.