The optimal rate of fluid administration to correct the volume deficit associated with diabetic ketoacidosis remains undefined. Thus, this study evaluates prospectively in adult patients without extreme volume deficit two regimens of therapy that differed exclusively in the rate of fluid infusion. In protocol 1 (12 patients), normal saline was infused at 1000 mL/h (approximately 14 mL/kg per hour) in the initial 4 hours and 500 mL/h (approximately 7 mL/kg per hour) during the subsequent 4 hours. In protocol 2 (11 patients), normal saline was infused at half the rates of protocol 1. For plasma bicarbonate level, the increment from admission levels at 2, 4, 8, 16, and 24 hours after admission was greater with protocol 2 at all times, attaining a statistically significant difference at 4 hours (3.7 vs 0.7 mmol/L) and at 24 hours (13.2 vs 8.4 mmol/L). These data support cautious restraint in the volume of fluid replacement used in the therapy of diabetic ketoacidosis, since prompt recovery, lack of any harmful effects, and a significant reduction in the overall cost of medical therapy are achieved with lower rates of saline infusion.