Objective: To determine the pharmacokinetic profile of ciprofloxacin 20 mg/kg per day (10 mg/kg administered intravenously 12 hourly) in paediatric patients with severe sepsis. Design: Open and prospective. Setting: Tertiary referral multi-disciplinary ICU. Patients: Twenty patients (two groups - group A: 3 months-1 year; group B 1-5 years). Interventions: Timed blood samples were taken for pharmacokinetics after the first dose (D0), as well as day 2 (D2) and then between days 6-8. Measurements and results: Ciprofloxacin serum levels were measured by high performance liquid chromatograghy. Demographic and clinical data and all adverse events were noted. Standard pharmacokinetic variables were calculated by non-compartmental methods. Peak concentrations (Cmax) for group A were D0 6.1±1.2 mg/l, D2 9.0±1.8 mg/l and D7 5.8±1.3 mg/l and, for group B, 7.4±1.3 mg/l, 7.8±1.6 mg/l and 6.4±1.3 mg/l, respectively, for the study periods. Concentration 12 h after the start of infusion (Cmin) for all periods were 0.2 mg/l or less. Areas under the curve (AUC, 12 h) were group A: 15.6±1.3, 19.2±1.63 and 14.1±1.4 mg/h per l, and group B: 15.9±1.3, 18.0±1.7 and 13.2±1.26 mg/h per l. One patient presenting with seizures, initially controlled, had another convulsion and a further patient developed seizures whilst on ciprofloxacin. Cmax in these patients were higher than the average Cmax. The convulsions of both patients were easily controlled. No other drug-related serious adverse events occurred. No arthropathy was noted. Three patients died of their underlying disease. Conclusions: There was no accumulation of drug even after 7 days of administration. Our Cmax and AUC were lower than that achieved in a similar adult pharmacokinetic study. To achieve end points of area under the inhibitory curve (AUIC) of 100-150 mg/h per l, 10 mg/kg ciprofloxacin eight hourly would be required for some resistant ICU organisms.