The optimal intravenous solution for rehydration of infants and children with severe dehydration is debated. AIM: The aim was to compare the efficacy of a polyelectrolyte solution (group PS) with sodium chloride 0.9% solution (group NS) in rapid parenteral rehydration of severely dehydrated infants with acute diarrhoea. METHODS: Primary outcomes were volume and time to hydration. Secondary outcomes were urea, creatinine, electrolytes, glucose, arterial pH and bicarbonate levels. Patients were assigned randomly and openly to one of the two treatment groups. Severe dehydration was defined as one or more of the following associated with any other sign of dehydration: depressed consciousness, a weak or absent pulse or capillary refill time > 10 sec. Peripheral blood samples for chemical pathology were collected before and after rapid fluid therapy. The mean age of the 36 enrolled infants was 9.1 mths. All had depressed consciousness or severe hypotension/shock. The fluid infusion rate was 50 ml/kg/hr until haemodynamic stability was restored (absence of severe hypotension and two urine emissions). Fluid volume, time to rehydration and weight before and after rehydration were recorded. RESULTS: All infants recovered full pulse within 1 hr; most had a better level of consciousness or capillary refill <3 sec. Group NS (15 infants) showed (before and after treatment, respectively) a decrease of plasma potassium (3.4 to 3.1 mmol/L, p=0.07), bicarbonate (13.3 to 12.2 mmol/L, p=0.01) and glucose (8.2 to 5.8 mmol/L, p<0.01). Group PS (21 infants) showed a decrease of potassium (4.4 to 3.2 mmol/L, p<0.01) but an increase in bicarbonate (11.6 to 13.3 mmol/L, p<0.01) and glucose (11.4 to 14.8 mmol/L, p=0.08). CONCLUSION: Polyelectrolyte solution was as effective as normal saline on volume expansion and better for correcting acidosis.