In theory the mortality rate from acute tubular necrosis should be low. The reasons for this are several. First, the damaged renal tubule is able to undergo more or less complete regeneration. Second, physiologic principles have been applied to minimize the fluid and electrolyte alterations which accompany acute renal failure. Third, there has been almost universal acceptance of hemodialysis as a procedure capable of improving many of the biochemical abnormalities and the symptoms attributable to them. Thus, Strauss and Raisz state, "The mortality in acute tubular necrosis itself can be reduced by the combination of physiologic measures and external dialysis to nearly zero." 1 Some reports give the impression that this goal is being achieved.2-5 Standing in contrast to this theory, however, are the discouraging results obtained at the renal centers with the most experience despite the best management currently available, including dialysis. Our mortality rate in 100 cases