The objective of this prospective study was to examine the effect of fluid restriction on body water and the outcome of children with acute meningitis. Fifty consecutively hospitalized children with acute meningitis, divided into two groups (A, without hyponatremia; and B, with hyponatremia), were randomly assigned to receive either normal maintenance (M) or restricted (R subgroup) (65 to 70% of M subgroup) fluids during the first 48 hours. Total body water, extracellular water (ECW), serum and urinary sodium and plasma and urinary osmolality were measured at admission and after 48 hours. In both groups children receiving restricted fluids showed a significant decrease in the mean total body water and ECW whereas body water remained unchanged in those on maintenance fluids. Children having an ECW reduction of 10 ml/kg or more in 48 hours had a significantly lower intact survival (10 of 28, 36%) than those with < 10 ml/kg or no reduction of ECW (15 of 22, 64%) (P < 0.05). The mortality was also higher in the former (7 of 28, 25%) than in the latter (2 of 22, 9%). On multiple stepwise regression analysis, ECW volume at admission (partial r2 0.20), ECW loss in 48 hours (partial r2 0.13) and plasma osmolality at admission (partial r2 0.22) were significantly related to outcome. We conclude that fluid restriction does not improve the outcome of acute meningitis. Indeed a decrease in ECW volume at 48 hours increases the likelihood of adverse outcome.