Erythropoietin (EPO) is recommended in individuals progressing toward end-stage renal disease (ESRD) to correct anemia and its complications, which are common in this disease. This study evaluated the impact of EPO administered before dialysis on mortality in incident ESRD patients. A total of 4,866 patients whose exposure to pre-ESRD EPO was determined from Health Care Financing Administration 2728 forms were analyzed. The median follow-up was 26.2 months, with 1,107 (22.7%) patients given EPO and 1,892 (38.9%) deaths. EPO use was more common in patients who had insurance before dialysis, remained employed, were started on renal replacement therapy outside the hospital, or initiated on peritoneal dialysis, which could be indicative of early intervention or quality care. The risk of death after starting dialysis was lower for patients treated with EPO before dialysis compared with patients who were not treated (adjusted relative risk 0.80, 95% confidence interval 0.70 to 0.91). There was no direct relationship between predialysis hematocrit and mortality; however, the most significant survival benefit with EPO use was in patients with the highest hematocrit values (adjusted relative risk 0.67, 95% confidence interval, 0.51 to 0.89). The most significant effect of pre-ESRD EPO use was observed during the first 19 months after starting dialysis (adjusted relative risk, 0.81; 95% confidence interval, 0.71 to 0.91), but this benefit diminished in patients with longer follow-up on renal replacement therapy. Use of EPO before dialysis confers a survival benefit to ESRD patients, especially in patients with an adequate hematocrit response before initiation of dialysis.