The aim of this study was to evaluate whether osteoprotegerin - an emerging inflammatory biomarker in cardiovascular diseases - predicts outcomes in patients with acute heart failure and preserved ejection fraction.We measured urea, creatinine, hemoglobin, high-sensitivity C-reactive protein, N-terminal pro-B-type natriuretic peptide and osteoprotegerin on admission in 177 patients admitted with decompensated heart failure and left ventricular ejection fraction at least 45%. The population was divided according to the median values of osteoprotegerin (158.6 ng/l). Primary and secondary endpoints were all-cause mortality and death/readmission at 1-year follow-up, respectively. Multivariable Cox models were generated for osteoprotegerin and common risk factors. We also evaluated the reclassification of patients into risk categories after adding this biomarker to the model.A total of 43 patients died during the follow-up and 84 had a combined event. Kaplan-Meier curves showed significantly increased primary and secondary endpoints according to the median of osteoprotegerin (log-rank, P < 0.0001 and 0.001, respectively). After adjustment for age, estimated glomerular filtration rate, hemoglobin, N-terminal pro-B-type natriuretic peptide, BMI and New York Heart Association III-IV, osteoprotegerin was a significant predictor of primary endpoint evaluated as continuous and categorized variable (relative risk 2.49, 95% confidence interval 1.18-5.24, P = 0.016 and relative risk 2.35, 95% confidence interval 1.11-4.96, P = 0.025, respectively). The clinical prediction model with osteoprotegerin evaluated with Net Reclassification Index was not significant.Osteoprotegerin is independently associated with all-cause mortality in patients hospitalized for heart failure with preserved ejection fraction. However, adding this biomarker into a risk model does not improve its prediction value.