Deuterium dilution for the measurement of total body water (TBW) has been conducted using varying protocols for equilibration. We measured TBW from deuterium dilution in urine samples in twenty-eight subjects using three protocols: (1) early morning dosage without breakfast, measuring deuterium in a second voiding at 4 h and 6 h; (2) early morning dosage with breakfast with the same measurement times; (3) dosage as last consumption before overnight sleep, measuring deuterium in a second voiding at 10 h. Results were compared with TBW estimates from underwater weighing (UWW). Because UWW is an indirect measure of TBW, it is used as an independent reference method in order to compare only relative discrepancies between the two methods. TBW values in the fasted state were not significantly different from those obtained in the fed state. The urinary deuterium enrichment was higher at 4 h than at 6 h (resulting TBW differences: 0.6 (SD 0.4) l). At 4 h and 6 h, differences in TBW measurements from deuterium and densitometry were positively related to the amount of TBW, indicating incomplete equilibration in larger water compartments. At 10 h no such relationship existed, indicating complete mixing of deuterium. It is concluded that 10 h equilibration is preferable to the shorter 4 h and 6 h, for the determination of TBW.