The objective of this study was to compare the pharmacokinetics of cefepime administered by continuous infusion and intermittent injection regimens. A prospective, randomized, cross-over study of ten patients with Gram-negative bacilli bacteraemia was conducted. All patients were randomized to receive cefepime either as a 4-g continuous infusion over 24 h for 48 h or a 2-g bolus administered intermittently intravenously every 12 h for 48 h. After 48 h the patients received the alternative dose regimen. Cefepime pharmacokinetic studies were carried out during hours 36-48 after the start of both regimens. All of the pathogens isolated from the blood in 7 patients had a minimum inhibitory concentration (MIC) < 1 microg mL(-1). In both regimens, the serum cefepime concentrations at all time points were higher than the MIC for the pathogens isolated from this study. For the continuous infusion arm, the highest steady-state concentration was 49.80+/-18.40 microg mL(-1) and the lowest steady-state concentration was 41.42+/-16.48 microg mL(-1). The steady-state concentrations were greater than 4 times the MIC of 8 microg mL(-1). For the intermittent injection regimen, the mean trough concentration was 4.74+/-3.99 microg mL(-1). The mean serum cefepime concentration was above 8 microg mL(-1) for 81.66% of the dosing interval. Therefore, we conclude that either continuous infusion or intermittent injection can be used as an effective mode of cefepime administration to achieve bactericidal activity.