The MICs and MBCs of benzylpenicillin, ampicillin, cefotaxime, and methicillin were evaluated against a Streptococcus (Enterococcus) faecium wild-type strain and against three mutants hyperproducing PBP 5 in cells incubated at both optimal and suboptimal temperatures. In the wild-type strain grown at optimal temperature, the MBCs of all beta-lactams were significantly greater than the MICs (bacteriostatic effect). As opposed to this, in the same cells grown at suboptimal temperature and in the mutants hyperproducing PBP 5 at all temperatures, the MICs of all antibiotics coincided with the MBCs (bactericidal effect). Under all conditions in which the MIC and MBC were the same, with all antibiotics, growth inhibition occurred only at the minimal concentration saturating all penicillin-binding proteins (PBPs) (or at higher concentrations). On the contrary, under conditions in which the MIC was lower than the MBC, only some of the PBPs were saturated (or bound) at both the MIC and the MBC, PBP 5 in no case being either saturated or bound. Under all conditions in which saturation of all PBPs was needed for growth inhibition, cells died at all antibiotic MBCs with kinetics which were much faster than those with which they died at the MBCs under conditions in which not all PBPs were saturated (or bound). In addition, under the former conditions, antibiotic concentrations above the MBCs did not significantly accelerate cell death kinetics, while under the latter conditions there was an acceleration in kinetics with increasing antibiotic concentrations up to full saturation of PBPs. It is suggested that the killing that occurs when all PBPs are saturated is a direct consequence of inactivation of PBP functions, while killing occurring when only some of them are saturated or bound is also (or mainly) an indirect consequence of inability of cells to grow and that, in S. faecium, the targets for growth inhibition and cell killing reside in different PBPs: for the latter effect, inactivation of one (or more) of the high-molecular-weight PBPs is sufficient, whereas in the former case inactivation of PBP 5 is necessary (after saturation of all other PBPs).