Glutamine-supplemented amino acid solutions are being utilized more frequently, especially following Food and Drug Administration approval and commercial availability of glutamine (Gln) dipeptides for parenteral infusion in Europe. It is important now, through evidence-based clinical investigation, to determine the appropriate patient groups who will benefit from this advancement. In the initial clinical outcome study, patients undergoing bone marrow transplantation were randomized to receive a Gin-supplemented parenteral mixture or standard Glnfree intravenous nutrition. I In the Gin-supplemented group, infection was decreased, length of stay was reduced, and hospital cost was significantly tess) The Gin-supplemented patients had a greater lymphocyte count at the time of bone marrow recovery, with T-cell subsets predominating. 3 These findings have been supported by other clinical investigations. Individuals receiving Gin-containing parenteral nutrition following bone marrow transplantation or major operation have a greater leukocytosis, 4 improved T-cell DNA synthesis, 5 the enhanced ability to express the anti-inflammatory cytokine interleukin-106, and the increased capacity to secrete the less toxic cystinyl-leukotrienes. 6 Gin administration has also been associated with decreased complications, improved nitrogen balance, and a reduced length of hospital stay. 4'6'7 In addition to these effects on immune function, Gln also supports the bowel, particularly by enhancing gut-barrier func, tion. 8 This clinical effect is most dramatically observed in premature infants who are born with an immature, leaky bowel. 9 Infants receiving Gin-fortified formulas reduced their bowelrelated infection rate from 30.2 to 11.4%. These and possibly other unknown effects of Gln may be important in interpreting the study by Griffiths et al. 1° published in this issue of Nutrition. This intensive-care team selected for study the most difficult group of patients cared for in their unit: those severely ill septic patients who required mechanical ventilation, couldn't be fed by the enteral route, and had a high incidence of multiple-organ failure. Only patients with liver failure and end-stage cancer (plus pregnant women and children) were excluded from this trial. Because of the heterogeneity and high mortality of this group, these patients are infrequently studied; yet it is this same patient group that requires new therapy in order to improve their outcome. One-half of the patients studied were randomized to receive a Gin-containing parenteral solution and the other half received conventional total parenteral nutrition. At entry, the groups were well matched as determined by multiple assessment criteria, including a variety of techniques to evaluate health status. As the study progressed, patient deaths occurred, initially from the underlying disease process but later from the consequences of sepsis and multiple-organ failure. Of significance, however, was the difference in late mortality observed in the two groups. In those patients fed > 10 d, 12/17 (71%) of controls died versus only 5/18 (28%)of the, Gin-fed patients (P = 0.03). At 6 mo, significantly more patients receiving Gln were alive than those who received conventional feedings (24 versus 14; P = 0.049). Why should this be so? First, we have already catalogued a number of biochemical mechanisms and physiologic functions that are influenced by Gin, and these effects may be highly beneficial to such critically ill patients. However, I would propose that Gin is playing a more global regulatory role by modifying the endogenous inflammatory response, possibly by attenuating the elaboration of proinflammatory mediators, and/or by upregulating anti-inflammatory factors. These effects could be mediated through the enhanced synthesis of the important intracellular antioxidant glutathione, which occurs with Gin administration, 11 but I suspect additional mechanisms are involved. To address such a question, one would need to harvest both serum and cells from shnilar patients so that circulating mediators and the cells' capacities to generate these signal proteins could be evaluated and the responses of these two patient groups compared. Blinded studies in acutely ill patients that extended for years (in this case 3 y, 9 mo) require unwavering faith in the scientific process. The unit must resist subtle changes in care that would jeopardize the study, and investigators must stay motivated and committed to the recruitment of patients and performing the details involved with such an investigation. All individuals involved with direct patient care must remain blinded.