One hundred consecutive low-birth-weight infants (less than 1751 g) were randomized equally into "dry" and "control" groups. Fluid administration in the dry group followed the schema 50, 60, 70, 80, 90, 100 and 120 ml/kg/day during the first week and 150 ml/kg/day until aged four weeks and that of the control group 80, 100, 120 and 150 ml/kg/day during the first week and 200 ml/kg afterwards. The diagnosis of bronchopulmonary dysplasia was based on typical roentgenographic findings, clinical symptoms and a supplemental oxygen requirement at age one month. Mean weight loss was greater in the dry group than in the control group (8.8 vs 6.2%), but there was no difference in weight between the groups at the age of four weeks. Twelve of the patients died, one in the dry group and 11 in the control group (five died during the first 24 h and the remaining six between three days and three months). Twenty-seven subjects in the dry group and 15 in the control group were alive and had no signs of bronchopulmonary dysplasia at age 28 days (p less than 0.05), the numbers at term being 28 and 14, respectively (p less than 0.01). These results suggest that fluid restrictions can reduce the mortality and morbidity of low-birth-weight infants. Infants of this kind, with respiratory distress, show poor tolerance to fluids, especially during the first weeks of life.