Acute renal failure research has been hampered by the lack of useful physiologic surrogate endpoints. Acute renal failure prevention and therapy studies using variables such as urine output and serum and urine chemistries have not yielded interventions proven to decrease the morbidity and mortality associated with acute renal dysfunction. Of those interventions that have been successful in smaller, phase II-level efficacy studies, none subsequently decreased the incidence of clinical (effectiveness) endpoints such as dialysis requirement or mortality in larger phase III trials. Suitable physiologic endpoints are needed to test the efficacy of new proposed therapies for the prevention and management of acute renal failure. Candidate endpoints for efficacy studies in acute renal failure prevention and management include glomerular filtration rate markers, renal blood flow, urine markers, and urine output. Possible endpoints for efficacy studies of renal replacement therapy in acute renal failure include serum markers of renal function and a variety of nonrenal markers. In this article, we present an approach to the choice of physiologic endpoints to determine the efficacy of interventions in acute renal failure.