An experimental Klebsiella pneumoniae pneumonia in rats was used to study the influence of continuous or of intermittent (8-hr intervals) administration of ceftazidime on therapeutic efficacy. Antimicrobial response was evaluated with respect to the calculated total daily dose that protected 50% of the animals from death (PD50) until 16 days after termination of a four-day treatment. When antibiotic treatment was started 5 hr after bacterial inoculation, the PD50 values after continuous and after intermittent administration of ceftazidime were 0.36 and 1.42 mg/kg per day, respectively (P less than .001). With a delay in the administration of the antibiotic to 34 hr after inoculation, the respective PD50 values were 1.08 and 13.06 mg of ceftazidime/kg per day (P less than .001). These studies show an improved therapeutic efficacy that increased with a delay in treatment when ceftazidime was administered by continuous infusion as compared with administration at 8-hr intervals. Continuous administration of PD50 doses of ceftazidime resulted in serum levels that were constantly below the MIC of the infecting Klebsiella strain.