Our current ability to favorably influence the adverse consequences of malnutrition in adult patients with established cancer is quite limited. Nutrient provision alone has not been successful in this regard. In fact, the approach of managing cancer patients with weight loss by solely providing calories is almost entirely extrapolated from clinical situations in which the presence of cancer is not a confounding problem and, therefore, may well be seriously flawed. These conclusions may not apply to situations in which special considerations hold, such as childhood malignancies and bone marrow transplantation. Current clinical management strategies for the cancer patient with weight loss require appropriate attention to the potential influence of the selected intervention on more than one parameter. As illustrated in Figure 1, nutritional support, whether by nutrient provision, pharmacologic administration, or a combination approach, differentially influences several parameters including nutritional status, abnormal host metabolism, gastrointestinal symptoms, and/or tumor growth. Changes in these parameters will influence the true end points with clinical relevance, which are patient survival and quality of life. Increased survival of patients with metastatic cancer has been difficult to achieve, even using chemotherapeutic regimens targeted directly at cancer growth. Similarly, nutritional support for patients with advanced cancer has not demonstrated improvement in this refractory parameter. Therefore, at the present time, clinicians must judge whether a nutritional support modality will favorably or unfavorably influence patient quality of life. This end point is of emerging importance in studies of nutritional support in cancer populations. Potential interrelationships among parameters influenced by nutritional support and their effect on clinically relevant end points are conceptually outlined in Figure 1. It is likely that concurrent attention to both optimal provision of nutrients and reversal of abnormal metabolism will be required if successful nutritional support approaches are to be described. Currently emerging clinical results provide some optimism for the future, but they do not unequivocally support the present routine application of any one particular nutrition support strategy for the medical patient with cancer.