In the effort to improve the long-term outcome in critically ill patients, the utilization of anabolic agents, such as human recombinant growth hormone, has been proposed in order to reduce catabolism and improve nutritional status. A recent multicentre study regarding the use of human recombinant growth hormone in intensive care unit patients showed an unexpected increase in the mortality rate in human recombinant growth hormone-treated patients. This finding is in contrast with previous literature data reporting either no differences or an even lower mortality rate with the administration of human recombinant growth hormone. This review evaluates the possible reasons for this dramatic difference in outcomes between the multicentre study and the existing literature. Articles dealing with human recombinant growth hormone administration either in intensive care unit patients (n= 26) or in postoperative patients (n= 16) have been reviewed. Our analysis suggests that the low caloric intake given to patients enrolled in the multicentre study might have been inadequate to compensate for the hypermetabolism of these patients, and could not support the prolonged and delayed administration of high doses of human recombinant growth hormone. Whether the beneficial metabolic effects of human recombinant growth hormone translate into better clinical outcomes deserves further investigation. In addition, the careful selection of patients to be treated, and close monitoring of both the adequacy of caloric support and modality of human recombinant growth hormone administration would favour the safety of human recombinant growth hormone utilization in critical care settings.