This prospective study was undertaken to systematically analyze the predictors of mortality in the elderly in a developing country. All elderly patients with ARF hospitalized at this tertiary care centre over 1 year were studied. Various predictors analyzed were hospital-acquired ARF, causative factors of ARF, preexisting hypertension and diabetes mellitus, severity of renal failure (initial and peak serum creatinine, need for dialysis), and complications of ARF: infection during the course of illness; serum albumin levels and critical illness defined as presence of two or more organ system failures excluding renal failure. Of 33,301 patients admitted, 4,255 (12.7%) were elderly. Of these 69 (1.6%) had ARF. On analysis of the whole group, both young and elderly, age >60 years had an independent predictor of mortality (odds ratio 5.6, P = 0.001). Forty-two of the 69 (60.9%) elderly ARF patients died. The mortality was significantly increased in those elderly with hospital-acquired ARF (79.2%, P = 0.027), those with sepsis as a cause of ARF (71.2%, P = 0.004), those who required dialysis (72.5%, P = 0.022), those developing an infection during the course of ARF (87.9%, P = 0.000) and in those with a critical illness (90.0%, P = 0.00). On logistic regression analysis of those variables that were significant on univariate analysis, only critical illness (odds ratio 9.97) and infection during course (odds ratio 9.72) were the independent predictors of mortality. To conclude, ARF complicates only 1.6% of hospitalized elderly patients but is associated with a high mortality rate of 61%. Infection during the course of illness and critical illness were the independent predictors of mortality.