OBJECTIVES: Cefepime (CPM) 2g q8h is FDA-approved for the empiric treatment of patients with febrile neutropenia. Pharmacokinetic-pharmacodynamic (PK-PD) studies have demonstrated that cephalosporin efficacy is associated with the percent-time that the free drug concentration exceeds the minimum inhibitory concentration (free %T>MIC). This study compares the free %T>MIC achieved by CPM 2g q8h to four alternative cefepime regimens in an attempt to determine whether antibiotic expenditures can be reduced without compromising predicted efficacy. METHODS: MIC distributions for Enterobacteriaceae, Pseudomonas aeruginosa, and Acinetobacter baumanii were extracted from the 2002 Intensive Care Unit Surveillance System database. A 10,000-subject Monte Carlo simulation was executed for each species/regimen pair using published PK parameters and these MIC distributions. Cefepime probabilities of target attainment (PTA) were determined for PK-PD targets of ≥50, 60, and 70 %T>MIC. PTA for CPM 2g q8h was compared to the PTA of four alternative regimens. RESULTS: Overall, the 2002 ISS database contained 3,543 Enterobacteriaceae (percent susceptible, 95%), 1,260 P. aeruginosa (53%), and 271 A. baumanii (36%) isolates. All CPM regimens achieved a PTA similar to that of CPM 2g q8h against Enterobacteriaceae suggesting an opportunity for cost minimization. In addition, CPM 1g q6h achieved equivalent or higher PTA than CPM 2g q8h against P. aeruginosa, and A. baumanii suggesting that the former regimen could be used to achieve similar predicted efficacy at a fraction of the daily cost (4g/d vs. 6g/d). However, the PTA vs. P. aeruginosa and A. baumanii was considerably lower for CPM 1g q8h, 1g q12h, and 2g q12h compared to CPM 2g q8h suggesting that these regimens may not be suitable alternatives for P. aeruginosa or A. baumanii. CONCLUSION: PK-PD models with Monte Carlo simulation demonstrate that cefepime 1g q6h results in similar predicted efficacy to cefepime 2g q8h despite a reduction in total daily drug of 2g/d.