At least 12 observational studies of large databases since 2002 (greater than 1.2 million patients) have reported a survival disadvantage with higher estimated glomerular filtration rate (eGFR) or renal clearance at the time of dialysis initiation. This study found the association between mortality and early dialysis initiation is affected by which equation is used to estimate glomerular filtration rate. The mortality risk was related to time of dialysis initiation, which is influenced by the GFR estimating equation. Specifically, results suggest that use of the 4-variable Modification of Diet in Renal Disease (MDRD) eGFR equation in patients with end stage renal disease may provide an inaccurate estimation of eGFR relative to the 6-variable MDRD equation. The increased risk for mortality from "early" dialysis initiation reported in some recent observational studies may be the result of misclassifying sicker, sarcopenic patients as early starts when the 4-variable MDRD equation is used.