Ceftazidime pharmacokinetics in 28 preterm infants (gestational ages, 25.6 to 31.9 weeks) were studied on day 3 of life. Patients with suspected septicemia were randomized on day 1 of life in two groups. One group (n = 13) was administered 25 mg of ceftazidime per kg of body weight once daily, and the other (n = 15) was given 25 mg of ceftazidime per kg twice daily. Both groups also received 25 mg of amoxicillin per kg twice daily. Blood samples were collected on day 3 of life with an arterial catheter at 0, 0.5, 1, 2, 4, 8, and 12 h after an intravenous bolus injection. An additional blood sample was taken at 24 h from the group dosed once a day. High-performance liquid chromatography was used to determine serum ceftazidime concentrations. The pharmacokinetics of ceftazidime were best described by using a one-compartment model. The half-life for the elimination of the drug from serum, apparent volume of distribution, total body clearance of ceftazidime, and inulin clearance were not significantly different for both groups. The ceftazidime/inulin clearance ratio was 0.72 for both groups. However, trough concentrations in serum for the twice-daily group were significantly (P < 0.001) higher (42.0 +/- 13.4 mg/liter) than those for the once-daily group (13.1 +/- 4.7 mg/liter). The latter concentrations were all still substantially higher than the MIC of ceftazidime for major neonatal pathogens. We conclude that the currently recommended dosage of 25 mg of ceftazidime per kg twice daily for preterm infants with gestational ages below 32 weeks may be adjusted during the first days of life to one daily dose at 25 mg/kg, provided that for the empirical treatment of septicemia, amoxicillin at 25 mg/kg is also given twice daily.