Objective To evaluate the impact of nutrition support on the clinical outcomes of gastrointestinal disease patients at nutritional risk and explore the the cost-effectiveness of various nutrition support options.Methods A prospective cohort study was designed.The patients who met the predetermined inclusion criteria were followed up during the hospital stay.Nutritional risk was determined using the Nutrition Risk Screening 2002 on admission.The information with respect to nutritional support,occurrence and treatment of complications,length of hospital stay,and discharge destination was monitored and recorded.The direct costs of nutritional support and the costs of diagnosing and treating complications were calculated after discharge.Infectious complication-free patient was used as the effectiveness indicator in the cost-effectiveness analysis.Results Patients who had received nutrition support had significantly lower infectious complications incidence (6.8％ vs.19.6％,x2 =9.0,P=0.003) and significantly higher total hospitalization costs (P =0.0001).The adjusted (by general linear model) cost of parenteral nutrition (PN) cohort,enteral nutrition (EN) cohort,PN combined EN cohort,and the cohort without nutritional support were 5635,1212,5220,and 1339 China Yuan,respectively.The incremental cost effectiveness ratios were 36 101,-794,and 33 748 China Yuan for PN,EN,and PN-EN combination groups,respectively.Conclusions For the patients at nutritional risk,nutritional support can remarkably reduce the incidence of infectious complications.The preliminary resuits of cost-effectiveness:due to lack of enough data required by health economic professional,it can not be cited directly.