Aim: To compare the efficacy of furosemide infusion with that of low-dose dopamine infusion in improving urine output and subsequent renal function in preeclamptic/eclamptic patients with oliguria in the immediate postpartum period. Design: Prospective randomised single blind clinical trial. Setting: Obstetric High Care Unit of King Edward VIII Hospital, a large referral tertiary hospital. Method: Eighty postpartum patients with severe preeclampsia/eclampsia with oliguria were enrolled. Hypovolaemia was corrected under central venous pressure (CVP) monitoring and urine output monitored for 4 hr. Patients who remained oliguric were randomly assigned to a continuous infusion of low-dose dopamine (3 μg/kg/min), or furosemide 5 mg/hr infusion, for 12 hr. In patients with no response after 12 hr, the drugs were switched and continued for a further 12 hr. A subgroup of patients who responded 4 hr after correction of hypovolaemia was observed for 12 hr. The primary outcome measured involved the comparison in urine output between the different drug regimes and the number of patients requiring haemodialysis. Secondary outcome measures involved assessment of serum urea and creatinine values in the two treatment groups. Results: Of the 80 patients enrolled, 20 improved their urine outputs within the 4-hr observation period. Sixty patients were randomised to furosemide or low-dose dopamine infusion. There was no statistical significant difference in the mean hourly urine output, rate of change in urine output over time and the mean urea or creatinine levels between the treatment groups. Ten percent of patients that failed on furosemide primarily, and 8.5% of patients that failed on initial low-dose dopamine, received haemodialysis. The difference in demographic and clinic data between these groups was not statistically significant. Conclusion: Administration of continuous infusion of furosemide showed comparable efficacy to low-dose dopamine infusion in ameliorating oliguria in severe preeclampsia/eclampsia post delivery; there was no difference in the percentage of patients that required haemodialysis in either group.