A mutant strain of Escherichia coli (E. coli ML-35) was used to follow the kinetics of phagocytosis, perforation of the bacterial cell envelope, and inactivation of bacterial proteins by human neutrophils. This particular E. coli mutant strain has no lactose permease, but constitutively forms the cytoplasmic enzyme beta-galactosidase. This implies that the artificial substrate ortho-nitrophenyl-beta-D- galactopyranoside cannot reach the beta-galactosidase unless the bacterial cell envelope has been perforated. Thus, the integrity of the E. coli envelope can be measured simply by the activity of beta- galactosidase with this substrate. Indeed, ingestion of E. coli ML-35 by human neutrophils was followed by perforation of the bacteria (increase in beta-galactosidase activity). Subsequently, the beta- galactosidase activity decreased due to inactivation of the enzyme. With a simple mathematical model and a curve-fitting computer program, we have determined the first-order rate constants for phagocytosis, perforation, and beta-galactosidase inactivation. With 32 normal donors, we found an interdonor variation in these rate constants of 20% to 30% (SD) and an assay variance of 5%. The perforation process closely correlated with the loss of colony-forming capacity of the bacteria. This new assay measures phagocytosis and killing in a fast, simple, and accurate way; it is not hindered by extracellular bacteria. Moreover, this method also measures the postkilling event of inactivation of a bacterial protein, which permits a better detection of neutrophils deficient in this function. The assay can also be used for screening neutrophil functions without the use of a computer program. A simple calculation suffices to detect neutrophil abnormalities. Neutrophils from patients with chronic granulomatous disease (CGD) showed an impaired rate of perforation and thus also of inactivation. Neutrophils from myeloperoxidase-deficient patients or from a patient with the Chediak-Higashi syndrome only showed a retarded inactivation of beta-galactosidase, but normal ingestion and perforation. The role of myeloperoxidase in the killing process is discussed. Although myeloperoxidase does not seem to be a prerequisite for perforation, it probably plays a role in bacterial destruction by normal cells, because the inactivation of bacterial proteins seems strictly myeloperoxidase dependent.