The luminol-dependent chemiluminescence of macrophages during the zymosan-stimulated respiratory burst has been studied both in the absence and in the presence of the radical inhibitor 3,5 di-tert-butyl-4-hydroxyphenyl propionic acid. In addition, the consumption of luminol and of the inhibitor has been followed analytically. Based on the rates of the consumption of the inhibitor, an iteration procedure yields a value of 2.2 x 10(-7) M for the steady-state concentration of radicals generated by cells at the maximum of the chemiluminescence in the presence of inhibitor. Approximate calculations have indicated that under the experimental conditions applied, additional formation of superoxide anion radicals by the oxidation of luminol is negligible. By assuming that in an inhibitor-free system the disappearance of radicals takes place via their combination process as well as by their interaction with luminol and/or with luminol-derived species, numerical integration yields a calculated curve of radical concentration versus time in fair agreement with experimental data and a rate-constant value for the combination of radicals of approximately 10(6) M-1 s-1, supporting literature findings according to which primarily superoxide anion radicals are formed.