SUMMARY The effect of bumetanide in infants with congenital heart disease presenting with cardiac failure was studied. The study was divided into acute (3 days) and long-term (mean 10 5 weeks) cases. A total of 12 male infants was included in the acute study and 13 cases were evaluated in the long-term study. The dose used in the acute study (0-015 mg/kg) was suboptimal; notwithstanding, it was found to cause significant natriuresis and chloruresis. Bumetanide in doses varying in different infants from as little as 0 015 mg/kg on alternate days to as much as 0 10 mg/kg daily was shown to be an effective diuretic for long-term use. No side effects were observed in either study.Bumetanide (Burinex) is a potent loop diuretic. Although it bears some structural resemblance to frusemide, unlike the latter which is derived from sulphanilamide, bumetanide is a metanilamide derivative (Feit, 1971). In dogs it is weight for weight 40 to 60 times more active than frusemide when given intravenously and 100 times more active when given orally (Ostergaard et al., 1972). In rats, bumetanide is practically ineffective. In human studies 1 mg bumetanide was equipotent to 40 mg frusemide with regard to urinary excretion of sodium, chloride, potassium, and water (Asbury et al., 1972). In healthy volunteers the major action of bumetanide has been shown to be on the ascending limb of Henle's loop with an additional effect on the proximal tubule but no significant effect on the distal tubule (Bourke et al., 1973). In this study we evaluated the effect of bumetanide in infants with congenital heart disease presenting with cardiac failure. The usual criteria for diagnosis of cardiac failure were used (Lees, 1969 diuretic was given but a preliminary 24-hour collection of urine was obtained. Only male infants were studied in order to facilitate accurate collection of urine samples. The 24-hour urine on day 1 was used as a control and was analysed for volume, sodium, potassium chloride, calcium, magnesium, phosphate, urate, glucose, and osmolality. On day 2 the infant was weighed and a blood sample was obtained and analysed for urea, sodium, potassium, chloride, calcium, magnesium, phosphate, liver function tests (serum proteins, SGOT, SGPT, alkaline phosphatase, bilirubin), haemoglobin, white cell count, and glucose. Immediately after blood sampling bumetanide 0 f015 mg/kg (prepared by dilution of the injectable form in sterile water) was administered orally. Urine was then collected at hourly intervals for 6 hours, followed by a further 18-hour collection to complete the 24-hour urine collection for day 2. On day 3 the infant's weight was recorded again and all the haematological and biochemical tests listed above were repeated.