Over the past two decades in the United States, there has been a inexorable increase in the mean estimated glomerular filtration rate (eGFR) at which patients are starting maintenance dialysis. Data from observational studies using eGFR derived from serum creatinine-based estimating equations has suggested that initiating dialysis at higher levels of kidney function may be associated with increased incidence of adverse clinical outcomes including mortality. At the same time, observational studies using time urinary clearances to measure eGFR have not demonstrated the same relationship, and instead of shown either no association between eGFR at dialysis initiation and subsequent outcomes, or increased survival with higher native renal clearances. There are numerous potential harms from the dialysis procedure itself, including adverse effects on cardiac structure and function, hemodynamic disturbances, increased oxidative stress, and risk for infection due to the presence of foreign objects such as dialysis catheters. Notwithstanding these potential harms, the IDEAL trial, the only controlled clinical trial published to date examining this question, showed no effect of eGFR on any of the examined clinically relevant outcomes at the start of maintenance dialysis. Thus, the available data do not support the use of renal function as the primary guide for determining timing of initiation of chronic renal replacement therapy. The nephrology community should transition to examining other potential approaches to addressing this important question. One potential area of interest is uremic symptomatology and health-related quality of life, though currently available data are indeterminate regarding which symptoms improve with the start of dialysis. Further research is needed to identify patient-centered approaches for decision making around dialysis initiation that will optimally improve survival and other important outcomes.