Continuous infusion (CI) of beta-lactam antibiotics provides a stable concentration which may result in a better activity against gram-negative bacteria if exceeding the minimum inhibitory concentration (MIC). Treatment outcome after 24 h CI of ceftazidime (CAZ) in cystic fibrosis (CF) children was compared with the bolus administration regimen. Fourteen CF children with chronic Pseudomonas aeruginosa pulmonary infection were treated during 14 days with the conventional CAZ thrice-a-day bolus infusion (regimen A), and few months later with 24 h Cl of CAZ (regimen B) using a portable pump. Amikacin was added to both regimens. Clinical efficacy of treatment was assessed using pulmonary, inflammatory and nutritional variables. Bacteriological analyses and CAZ concentrations in serum and sputum were also measured. All patients improved clinically with both regimens. Among the parameters used to compare both regimens, only prealbumin values improved (regimen A: + 0.08 g/l versus regimen B: +0.11 g/l P = 0.015). No clinically significant side-effects were noted. In regimen A, the mean predose (trough level) CAZ concentration in serum was highly variable (range 2.2-45.4 gg/ml) with some values (32% of samples) below the MIC of P. aeruginosa isolates found in the sputum of the patients. In regimen B, the serum CAZ level achieved was 28.5+/-8.4 microg/ml without any value below the MIC. The mean sputum levels were comparable in both regimens. No CAZ resistant strains of P. aerugino.sa appeared between and directly after the treatments. CONCLUSION: The clinical outcome of children with cystic fibrosis treated with 24 h continuous infusion of ceftazidime was no different from that achieved with the conventional bolus infusion regimen. Continuous infusion provided a sustained serum ceftazidime level well above the P. aeruginosa minimum inhibitory concentration. Continuous infusion was well tolerated and appreciated by the children and this may promote home therapy for cystic fibrosis children.