This study was designed to determine the need for water supplementation to maintain water homoeostasis in exclusively breastfed infants during summer in a tropical country. A prestudy questionnaire revealed that 97% of 34 nurses and 63% of 70 doctors advocated such supplementation. 45 healthy, male, exclusively breastfed babies, aged 1-4 months, were recruited from a well-baby clinic. 9 who had never received supplemental water plus a random selection of 14 others were allocated to group I (breastmilk only); the remaining 22 infants were allocated to group II (breastmilk plus supplemental fluid according to the mother's usual practice). The babies were studied at the hospital for 8 h; breastmilk intake was measured by weighing the infant before and after each feed, water intake by calibrated bottles, and urine output by accurate collection and measurement. The maximum room temperatures were 34-41 degrees C and relative humidities 9-60% (below 50% in all but 3 infants). In group II the mean water intake was 11% (95% confidence interval 7-16%) of the total fluid intake. Both breastmilk intake (274 vs 210 ml) and total fluid intake (274 vs 233 ml) were higher in group I than in group II (p = 0.003, p = 0.073, respectively), after adjustment for age, weight, length, room temperature, and humidity. However, there were no significant differences between the groups in urine output, urine or serum osmolality, weight change, or rectal temperature whether or not the factors adjusted for included total fluid intake. Thus, exclusively breastfed infants do not need supplemental water to maintain water homoeostasis; a reduced breastmilk intake is a potential disadvantage of this practice.