Recent advances in technology have not substantially changed the relatively low survival rate associated with acute renal failure (ARF). Several clinical prognostic variables and multivariate models have been reported to predict survival in individual patients, but these are either cumbersome to use or restrictive in their application. A straightforward clinical index has been developed to predict survival in ARF based on data obtained for all patients receiving dialysis for ARF at the University of Kansas Medical Center from November 1979 through October 1985. During this period, 126 patients received dialysis for ARF, with an overall survival of 25% (32/126). There were no significant differences between survivors and nonsurvivors in age, gender, or indication for dialysis. Eleven variables were statistically related to survival, and were reduced to five when clinically similar variables were combined or eliminated. A clinical survival index was based on these five easily determined variables that were significantly related to survival: systolic blood pressure less than or equal to 110 mm Hg, assisted ventilation, congestive heart failure, proven or suspected sepsis, and gastrointestinal (GI) dysfunction (bleeding, ileus, obstruction, or recent abdominal surgery). Survival was directly related to the number of factors present: zero, 62% (8/13); one, 44% (8/18); two, 30% (10/33); three, 19% (5/26); four, 0% (0/20); and five, 6% (1/16). This straightforward index, derived from easily obtained clinical data, is useful for judging survival prognosis in patients with ARF severe enough to warrant treatment with dialysis.