Study Objective. To compare the pharmacodynamics of meropenem and imipenem, both administered as 500 mg every 6 hours, against populations of Enterobacteriaceae, Acinetobacter baumannii, and Pseudomonas aeruginosa. Design. Ten thousand-subject Monte Carlo simulation. Intervention. Variability in total body clearance (Cl-T), volume of distribution as calculated by the terminal elimination rate (Vd(beta)), and minimum inhibitory concentration (MIC) distributions (Escherichia coli, Klebsiella pneumoniae, Enterobacter cloacae, Serratia marcescens, A. baumannii, R aeruginosa) were derived from the literature for both meropenem and imipenem. For the free drug concentrations, the percentage of the dosing interval that the drug concentrations remain above the MIC (%T>MIC) for each carbapenem-bacteria combination was calculated for 10,000 iterations, substituting a different Cl-T, Vd(beta), fraction of unbound drug, and MIC into the equation each time based on the probability distribution for each parameter. Probabilities of attaining targets of 30%, 50%, and 100% T>MIC were calculated. Measurements and Main Results. Meropenem free drug %T>MIC exposure was significantly greater than that of imipenem against Enterobacteriaceae and P. aeruginosa, whereas imipenem exposure was greater for A. baumannii. For both agents, free drug %T>MIC exposure was greatest against Enterobacteriaceae and less for A. baumannii and P aeruginosa. Probabilities of target attainment for 30% and 50% T>MIC were similar between drugs for most bacteria. At 100% T>MIC, meropenem target attainments were greater than those of imipenem against Enterobacteriaceae and P aeruginosa, and imipenem attainment was higher for A. baumannii. Conclusion. The probability of attaining lower pharmacodynamic targets for most gram-negative bacteria is similar for these carbapenems; however, differences become apparent as the pharmacodynamic requirement increases. Further study of the benefits of achieving this pharmacodynamic breakpoint with a higher probability of attaining targets is necessary.