Despite technical advances in the delivery of hemodialysis over the past decade, the mortality rate of hemodialysis-dependent, end-stage renal disease (ESRD) patients in the United States remains high. The increase in the number and severity of comorbid conditions of patients entering ESRD is a factor contributing to this high mortality. Nevertheless, there is increasing evidence that the dose of dialysis received by US patients is inadequate and that this plays a major role in the observed high mortality. In this review, we examine some of the parameters used to judge the adequacy of dialysis, as well as factors that can result in differences between prescribed and delivered dose of hemodialysis. Based on available evidence, we propose that for most patients the optimum dose of dialysis, above which further improvement of morbidity and mortality is doubtful, is represented by a delivered dose of dialysis equivalent to a Kt/V of 1.4 or greater, using biocompatible membranes. The prescription of this optimal dose of dialysis must be coupled with an ongoing effort to monitor delivery of the appropriate dose.