These preliminary studies indicate that empirical methods of controlling fluid balance of a patient during extracorporeal dialysis are fraught with many hazardous unknowns. Changes in plasma volume, red blood cell volume, and total blood volume during dialysis were determined by one group of workers, while another group performed the dialysis and regulated hydration empirically, as is customary. Actual measurements were later compared with the clinical estimates of the patient's needs. Marked disparities were found to exist. A firm basis for satisfactory regulation of hydration during dialysis would be afforded by the routine incorporation of blood volume studies during extracorporeal dialysis.