OBJECTIVE： To provide reference for strengthening the rational nutritional intervention of patients with malignant tumor. METHODS：From Sept. 2016 to Oct. 2017， 191 inpatients diagnosed as malignant tumor in our hospital were targeted. Nutritional support pharmacist used Nutritional Risk Screening 2002 （NRS 2002） and Patient-generated Subjective Global Assessment （PG-SGA） to conduct nutritional status assessment within 48 h of admission； at the same time， nutritional intervention for patients during hospitalization was investigated. RESULTS： Among 191 patients， NRS 2002 score ≥3 （with nutritional risk） was classified in 123 cases， and total incidence of nutritional risk was 64.40％. Totally 129 cases were classified into grade B and C （malnutrition） by PG-SGA， and total incidence of malnutrition was 67.54％. There was statistical significance in the incidence of malnutrition among patients aged below 60 years old and patients aged more than 60 years old （P＜0.05）. NRS 2002 and PG-SGA showed a consistency of Kappa＝0.557 （P＜0.001）， with moderate consistency degree. Of 191 patients， only 46 received nutritional intervention （only 2 cases received total enteral nutrition）， and the overall intervention rate was 24.08％. Among them， the intervention rate of patients with nutritional risk was only 34.15％ （42/123）， and the intervention rate of patients with malnutrition was 31.78％ （41/129）. CONCLUSIONS： The incidence of nutritional risk and malnutrition among patients with malignant tumors is high， especially in elderly patients. But clinical nutrition interventions are obviously inadequate and the application rate of enteral nutrition support is low. Therefore， nutritional support pharmacists are required to assist and cooperate with physicians， combined two tools as NRS 2002 and PG-SGA that can promptly and effectively screen patients who already have nutritional risk or malnutrition on admission and comprehensively evaluate their nutritional status， so as to give appropriate nutritional intervention.