The aim of this study is to evaluate the clinical outcome. Between January 1,2002 to September 30, 2002, we prospectively and retrospectively recruited III hospitalized patients who received Enteral Nutrition (EN group n = 52) and Total Parenteral Nutrition (TPNgroup n = 59) for more than seven days. The factors of clinical outcomes are costs, incidences of infection, lengths of hospital stay, and changes in weight. The characteristics of patients were investigated, which included nutritional status, disease severity CAP ACHE III score) and hypermetabolic severity Chypermetabolic score). Hypermeta-bolic scores were determined by high fever (> $38^{circ}C$), rapid breathing (> 30 breaths/min) , rapid pulse rate (> 100 beats/min), leukocytosis (WBC > 12000 $mm^{3}$), leukocytopenia (WBC > 3000 $mm^{3}$), status of infection, inflammatory bowel disease, surgery and trauma. There was a positive correlation between hypermetabolic score and length of hospital stay (ICU), medical cost, weight loss, antibiotics adjusted by age while APACHEIII score did not show correlation to clinical outcome. Medical cost was higher by $18.2\%$ in the TPN group than the EN group. In conclusion, there was a strong negative correlation between the clinical outcome (cost, incidence of infection, hospital stay) and hypermetabolic score. Higher metabolic stress caused more malnutrition and complications. For nutritional management of patients with malnutrition, multiple factors, including nutritional assessment, and evaluation of hypermetabolic severity are needed to provide nutritional support for critically ill patients.