Introduction: Hospital malnutrition (HM) is an increasingly prevalent situation, which involves both an increase in health costs, and also a decrease in the life quality and greater morbimortality. Nutritional screening is essential to detect malnutrition early and avoid these complications. Objectives: To develop, validate and implement the Nutritional Risk Screening 2002 (NRS-2002) tool at the admission and during the hospitalization of a patient in the oncohematology service in a third level hospital, and know their nutritional status at the admission and the discharge. Methods: NRS-2002 was performed on all patients admitted to the oncohaematology service, followed by the complete nutritional assessment (NA) to check its validity. NRS-2002 is repeated weekly to determine the degree of malnutrition during the hospital stay. Results: 573 patients were admitted to the oncohematology service, of which a 34.4% suffered from malnutrition, 44.7% had risk of malnutrition and 20.9% were in good nutritional condition, at the time they were admitted to hospital according to the NRS-2002. In patients admitted for more than a week, NRS-2002 was performed weekly and found that, upon discharge, a 34.4% were malnourished, 50.8% had a risk of malnutrition and the last 14.76% were in good nutritional status; also a 12.3% worsened their nutritional status, the 68.9% maintained it and only a 18.9% improved it. 78.8% of patients with longer admissions require a NA. Discussion: Due to the high risk of malnutrition in hospital admission, the use of nutritional screening is necessary, both at admission and during hospital stay to avoid nutritional deterioration during the same. Conclusions: Our results suggest that the NRS-2002 is a simple and effective method for early malnutrition detection.