To demonstrate the impact of normal (NEK) and impaired elimination kinetics (IEK) of gatifloxacin on its ability to protect from losses in the susceptibility of Staphylococcus aureus, a clinical isolate of methicillin-resistant S. aureus at a starting inoculum of 10(8) cfu/ml was exposed to 3 days of quinolone dosing. A series of monoexponential pharmacokinetic profiles with half-lives of 7 h (NEK) and 31 h (IEK) were simulated over 32- and 8-fold ranges of the 24 h area under the concentration-time curve (AUC(24))-to-MIC ratio, respectively. The simulated AUC(24)/MICS were designed to provide peak concentrations (C(max)s) close to the MIC, between the MIC and the mutant prevention concentration (MPC), i.e., within the mutant selection window (MSW), and above the MPC. With both NEK and IEK simulations, significant increases in MIC were observed at those AUC24/MICS that correspond to gatifloxacin concentrations within the MSW over most of the dosing interval (>25%). No such increases were observed at the smallest AUC(24)/MIC (10 h with NEK and 20 h with IEK) when the simulated C(max)s were close to the MIC, with minimal if any bacterial killing, and at the highest AUC(24)/MICS (310 and 160 h, respectively) when gatifloxacin concentrations exceeded the MPC over most of the dosing interval, with maximal antimicrobial effect. These 'protective' AUC(24)/MIC ratios correspond to 135% of the usual gatifloxacin clinical dose (400 mg IEK) and 60% of the loading and maintenance doses (400 mg, then 200 mg IEK). This study predicts different protective potentials of gatifloxacin in IEK and NEK against staphylococcal resistance and supports the MSW concept. (C) 2003 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.