OBJECTIVES: We wanted to establish optimum protein and glucose intakes during total parenteral nutrition by using a constant caloric but changing protein intake in critically ill, ventilated, anuric patients on continuous renal replacement therapy and measuring amino acid and glucose losses across the hemofilter. METHODS: Eleven consecutive, critically ill patients (eight male, age, 43.5 +/- 21.8 y; Acute Physiology and Chronic Health Evaluation II score, 20.5 +/- 7.0; Acute Physiology and Chronic Health Evaluation risk of death: 36.5% +/- 23.0 and 6 +/- 1 impaired organ systems) entered this study. Patients were fed by continuous infusion of a total parenteral mixture consisting of Synthamin (a mixture of essential and non-essential amino acids), 50% dextrose, and intralipid (long-chain triglycerides) to meet caloric requirements as predicted by Schofield's equation corrected by stress factors. The amount of protein infused was varied (1 to 2.5 g. kg(-1). d(-1)) by increments of 0.25 g. kg(-1). d(-1). Patients were stabilized on each feeding regimen for at least 24 h before paired samples of blood and dialysate were taken for amino acid and glucose measurements. Continuous renal replacement therapy was performed by using a blood pump with a blood flow of 100 to 175 mL/min. Dialysate was pumped in and out counter-currently to the blood flow at 2 L/h. A biocompatible polyacrylonitrile hemofilter was used in all cases. RESULTS: With protein intakes below 2.5 g. kg(-1). d(-1), blood levels of 14% to 57% of the measured amino acids were below the lower limits of the normal range. At 2.5 g. kg(-1). d(-1), all measured amino acids were within the normal range. Amino acid balance became more positive as protein input increased (P = 0.0001). Glucose and amino acid losses were dependent on blood concentration. Overall, 17% (range, 13% to 24%) of infused amino acids and 4% of infused glucose were lost in the dialysate. CONCLUSIONS: This study of critically ill, ventilated, anuric patients on continuous renal replacement therapy suggested that increases in protein and glucose are required to account for the increased losses across the hemofilter. A protein intake of 2.5 g. kg(-1). d(-1) appeared to optimize nitrogen balance and correct amino acid deficiencies.