Mineral excretion following single doses of furosemide were compared with bumetanide in a random cross-over trial in 17 premature infants. The mean birthweight and gestational age were 889 +/- 85 g and 27 +/- 2 weeks. Following furosemide therapy, significantly higher chloride losses and urine volumes were noted in the first 8-h period compared with the second or third 8-h periods. Following bumetanide therapy, sodium, calcium, and chloride losses and urine volumes were significantly higher in the first 8 h compared with the second or third 8-h periods. Hourly sodium and chloride losses were significantly lower following bumetanide than furosemide during the first two 8-h periods. During the final 8-h period sodium, potassium, chloride, and calcium losses were significantly lower following bumetanide than following furosemide. Sodium loss per urine volume was lower with bumetanide than furosemide but calcium loss tended to be higher. Hence, bumetanide does not appear to be a calcium-sparing diuretic following single-dose therapy.