In vitro, antibiotics are known to kill bacteria in predictable relationships to their broth concentrations. It was our hypothesis that serum concentration was an important determinant of the rate of bacterial eradication in patients. This contention was reinforced by animal studies, which have clearly demonstrated concentration-related antibacterial activity. The animal models rely on reduction of bacterial colony counts as an endpoint of effect, and demonstrate that colony count reductions are related to antibiotic dose and probably to serum concentration. We adapted these methods for use in intubated patients with Gram-negative pneumonia. Briefly, each patient had extensive staging of the pneumonic process, and the Gram-negative organism was isolated and its minimal inhibitory concentration (MIC) was determined. In each patient measurement of the antibiotic serum concentrations in the interval between two doses of the drug was also performed. The pharmacokinetic profile of the drug was then superimposed on the bacterial MIC, and we then derived the patients individual peak to MIC ratio, area above MIC, and time above MIC. Each of these pharmacokinetic parameters was then related to the time required to eradicate the bacterial pathogens in the patient. For beta-lactams and quinolones, time above MIC was the most predictive of bacterial eradication times. Clearly, these methods can be used to develop dosing strategies for patients, as well as to determine clinically relevant doses and dosing strategies in clinical trials. Further work is needed, however, to assess whether these concepts hold for other types of bacteria (such as Gram-positive) or apply as accurately to other infection sites in addition to pneumonia.