Multiple factors still influence the high rate of mortality in acute tubular necrosis. Trying to analyze the influence of each risk factor present in an individual patient and the possible interdependence between these factors, as well as to obtain an early prognosis, we have applied a forward analysis to demographic data, acute renal failure origin, need of dialysis, diuresis and clinical conditions in 228 patients, using a multiple linear regression model contained in a computer package. Based on this approach we have found that three variables: deep neurological coma, persistent blood hypotension and assisted respiration have significant influence on mortality. Also, a regression equation was obtained which could be applied as a discriminant score to patient prognosis. This score, calculated with the three aforementioned variables and oliguria when the nephrologist sees the patient for the first time, allows an easy and early prognosis in each patient with acute tubular necrosis.