The bactericidal activity of beta-lactams is time-dependent, and the time spent above the MIC (T > MIC) is the best predictor of efficacy. A prospective, randomized, open-label study was conducted in intensive care unit (ICU) patients with gram-negative rod infections to compare the efficacy of cefepime given as a continuous versus an intermittent infusion. Of the 18 patients included to date, 14 had severe pneumonia and four bacteremia. All patients received amikacin, 15 mg/kg/d, and cefepime, 4 g/d. Patients were randomized to cefepime administration as a continuous infusion (Group 1, n = 9) or as an intermittent infusion (Group 2, n = 9, 2 g every 12 h). No significant differences were found between the two groups for age, sex, initial infection, IGS II score (46 vs 48, NS) or the MIC of the gram-negative organism. Mechanical ventilation and hospital stay durations, recovery rates, and pharmacokinetic parameters (24-h AUIC, 12-h AUIC, T > MIC, and T > 5 x MIC) were compared in the two groups using the chi-square and Mann-Whitney tests. P values < 0.05 were considered statistically significant. There were no significant differences for mechanical ventilation duration, recovery rate, hospital stay duration (34 vs 36 days, NS), 24-h AUIC (624 vs 473, NS), or the 12-h AUIC (235 vs 238, NS). There were two interesting findings: T > MIC was significantly (P < 0.05) higher in Group 1 (23.84 +/- 0.2) than in Group 2 (20.7 +/- 3), and T > 5 x MIC was also significantly (P < 0.01) higher in Group 1 (23.61 +/- 0.6) than in Group 2 (16.6 +/- 6). Although clinical outcomes were similar in the two groups, it is reasonable to assume that the longer time spent with a cefepime level above the MIC in the continuous infusion group was associated with a more stable bactericidal effect.