The high mortality rate among dialysis patients has spawned investigation into potentially correctable factors that are associated with an increased risk of death. Several studies have demonstrated a strong association between an increased risk of death in dialysis patients and suboptimal delivered dose of dialysis, malnutrition, and non-renal comorbidity. In addition, the use of unsubstituted cellulose dialyzers and reprocessed dialyzers also has been associated with an increased risk of death. Increased attention to these factors has resulted in a significant improvement in patient survival. Nonetheless, the mortality of dialysis patients remains unacceptably high and indicates that other factors may be operative. One of the factors that has thus far received scant attention, but could significantly affect morbidity and mortality in dialysis patients, is the timing and quality of care before initiation of dialysis. Optimal pre-end-stage renal disease care involves early interventions aimed at delaying progression of chronic renal failure, judicious management of uremic complications, timely placement of vascular access, timely initiation of renal replacement therapy, and implementation of educational programs targeted at maximum rehabilitation. Given the fact that early referral to the nephrologist is likely to result in optimal pre-dialysis care, the 1993 National Institutes of Health Consensus Statement on Morbidity and Mortality of Dialysis recommended that referral of a patient to a renal team should occur at a serum creatinine of 1.5 mg/dL in women and 2.0 mg/dL in men. Several investigators also have argued that patients with chronic renal failure who begin dialysis at a relatively "high level of residual renal function" (early start) may have lower morbidity and mortality compared with patients who begin dialysis at a more traditional "low level of renal function" (late start). This hypothesis is based on evidence that declining renal function is associated with malnutrition and that malnutrition at the start of dialysis is associated with poor clinical outcomes. Furthermore, patients are started on dialysis at an endogenous solute clearance that is lower than that accepted as optimum for patients on dialysis. Finally, limited clinical studies have demonstrated the benefit of early initiation of dialysis. Consequently, the Peritoneal Dialysis Adequacy Work Group of the National Kidney Foundation-Dialysis Outcomes Quality Initiative recommends that dialysis be initiated when the weekly renal Kt/Vurea decreases to below 2.0 unless all three of the following criteria are fulfilled: (1) stable or increased edema-free body weight, (2) normalized protein equivalent of total nitrogen appearance greater than 0.8, and (3) absence of clinical symptoms and signs attributable to uremia.