The understanding of fluid fluxes in pediatric cardiac critical care is crucial to effective management. Knowledge of variations in total body water in this situation would aid this understanding, but most available methods are unsuitable for routine use. Recently, estimation of total body water by a tetrapolar bioelectric impedance has been validated in older children and adolescents. We undertook a study to validate the method in the taxing conditions of pediatric cardiac critical care. A prospective comparative study was done in 16 children whose ages ranged from 6 days to 10 years (mean 23 months) after a variety of cardiac operations. Total body water was estimated by a standard isotope dilution method (deuterium oxide) and by bioelectric impedance by means of a Holtain body composition analyzer. Individual estimations of total body water were made on two successive days on each patient at varying intervals after a cardiac operation, bioelectric impedance being measured hourly during 4 hourly urine collections for the deuterium oxide method. Thirty-two simultaneous values of total body water (by isotope and by impedance) were collected. Population-specific regression relationship was established by plotting total body water (isotope) against height2/bioelectric impedance. From this data plot r = 0.911, giving this equation: total body water = 0.158 +/- 0.662 x (height2/bioelectric impedance). Levels of agreement of -1.771 to +1.725 were observed, with a standard error of measurement of 16% across the range. The data suggest that bioelectric impedance is a satisfactory and reliable method of estimating total body water in children requiring cardiac critical care. The standard error of 16% suggests that the method may be more useful for measuring trends than absolute values, but the technique should be a valuable noninvasive tool both for continuous monitoring of total body water and in longitudinal research studies of rapid fluid flux and in the assessment of capillary leak.