Increase in mortality in hemodialysis patients in the United States is a rising concern. In this study, the annual mortality rate from 1981 to 1989 was determined and the morbidity and mortality of 136 patients dialyzed during 1988 was analyzed. All patients were followed with urea kinetics (UK). Patients were divided according to NCDS mapping of BUN and protein catabolic rate (pcr): Group 1, inadequate protein intake (n = 28); Group 2, adequate dialysis (n = 51); Group 3, excessive dialysis (n = 28); Group 4, undefined domain (n = 13); and Group 5, transitional domain (n = 16). The UK parameters measured were midweek predialysis BUN, pcr, and Kt/V. Group 1 by definition had the lowest pcr (0.7 g/kg/day vs. greater than 1.0 in the other groups). Patients in Group 1 had significantly lower BUN, creatinine, hematocrit, and serum albumin. Results of all other laboratory tests were not significantly different. Numbers of hospitalizations and days in the hospital were higher in Group 1. Causes of hospitalization were similar in all groups. Mortality rate was 8% in Group 1 vs. 4.4% in Group 2. Dialysis prescription using UK has maintained the annual mortality rate at about 14% for the past nine years.