The intent-to-treat analyses of all patients in the HEMO trial suggested that increases in dose of dialysis as measured by urea Kt/V were of marginal or no benefit when dialysis was provided in a 3 times/wk schedule. The as-treated analysis in the HEMO trial pointed to markedly increased mortality when the delivered dose decreased even slightly below the targeted dose, evidence of a dose-targeting bias. The intent-to-treat HEMO study results suggested a potential interaction between sex and the dose-mortality relationship, and this also has been found in some cross-sectional studies, the cause of which remains unexplained. Whether dialysis dose should continue to be targeted based on urea distribution volume (V), or targeted to a body size measure that is a lower power of body weight (such as body surface area), remains an open question. The lack of benefit of increasing the dialysis dose in a 3 times/wk setting is more understandable if one looks at measures of equivalent continuous solute removal, such as the standard Kt/V. Differences in standard Kt/V in the 2 dose arms of the HEMO trial, for example, were only about 15%. Without going into removal of very large solutes (eg, beta-2-microglobulin), which is discussed elsewhere in this issue, or protein-bound uremic solutes, the only way to provide significantly more dialysis dose may be to move to more frequent dialysis schedules and/or to very long session lengths. Here, benefit may be related as much to better control of salt and water balance as to better removal of uremic toxins.