Considerable information on the pharmacodynamics of beta-lactams has accumulated in the past 20 years. In vitro, beta-lactams demonstrate time-dependent killing and variable postantibiotic effects. Animal models have shown that the time for which drug levels exceed the minimum inhibitory concentration (MIC) correlates best with bacterial eradication, and this is now being borne out in human studies. In investigations on osteomyelitis and endocarditis, trough serum inhibitory titers have generally correlated better with cure than have peak titers, and studies that have analyzed outcomes in relation to the MIC for the infecting pathogen have shown decreasing clinical efficacy with increasing MICs. One prospective study has shown that time above MIC correlated better with time to pathogen eradication than did area under the curve. In some continuous-infusion studies, significantly better outcomes were achieved with continuous infusion against susceptible bacteria or for patients with persistent, profound neutropenia. With use of time above MIC as the predictor of efficacy, it is possible to reexamine current dosing schedules critically.