Background: Frusemide is frequently administered to critically ill patients in the intensive care unit (ICU). We investigated whether continuous frusemide infusion therapy was more effective than regular intermittent bolus doses at causing diuresis. Methods: 59 adult patients with fluid overload admitted to two tertiary multidisciplinary ICUs were randomised to either a continuous frusemide infusion or regular intermittent intravenous boluses of frusemide according to pre-defined algorithms aiming for a minimum hourly urine output. Results: There was no significant difference in diuretic response between the two groups during the first 24 h (5.3 liters in the bolus group vs. 5.4 liters in the infusion group). In the bolus group a significantly higher dose of frusemide was needed to achieve target diuresis (24.1 vs. 9.2 mg/h in the infusion group, p = 0.0002). Mean urine output per dose of frusemide was significantly higher in the infusion group (31.6 vs. 18 ml/mg in the bolus group, p = 0.014). At the end of the study, there were no differences in hospital mortality, number of patients requiring ventilatory support, change in serum creatinine or change in estimated glomerular filtration rate. Conclusions: Both intermittent boluses and continuous infusion of frusemide were successful in achieving algorithm-driven diuresis. However, continuous infusion therapy was more effective than intermittent boluses since the dose of frusemide required was significantly less.