SummaryControlling nutritional status (CONUT) uses 2 biochemical parameters (serum albumin and cholesterol level), and 1 immune parameter (total lymphocyte count) to assess nutritional status. This study examined if CONUT could predict the short-term prognosis of heart failure (HF) patients.A total of 482 (57.5%) HF patients from the Ibaraki Cardiovascular Assessment Study-HF (n = 838) were enrolled (298 men, 71.7 ± 13.6 years). Blood samples were collected at admission, and nutritional status was assessed using CONUT. CONUT scores were defined as follows: 0-1, normal; 2-4, light; 5-8, moderate; and 9-12, severe degree of undernutrition. Accordingly, 352 (73%) patients had light-to-severe nutritional disturbances. The logarithmically transformed plasma brain natriuretic peptide (log BNP) concentration was significantly higher in the moderate-severe nutritional disturbance group (2.92 ± 0.42) compared to the normal group (2.72 ± 0.45, P < 0.01). CONUT scores were significantly higher in the in-hospital death patients [4 (3-8), n = 14] compared with patients who were discharged following symptom alleviation [3 (1-5), n = 446, P < 0.05]. With the exception of transferred HF patients (n = 22), logistic regression analysis that incorporated the CONUT score and the log BNP, showed that a higher CONUT score (P = 0.019) and higher log BNP (P = 0.009) were predictors of in-hospital death, and the median duration of hospital stay was 20 days.Our results demonstrate the usefulness of CONUT scores as predictors of short-term prognosis in hospitalized HF patients.(Int Heart J Advance Publication)