AIMS: To describe the prevalence, overlap, and prognostic implications of physical and social frailties and cognitive dysfunction in hospitalized elderly patients with heart failure. METHODS AND RESULTS: The FRAGILE-HF study was a prospective multicentre cohort study enrolling consecutive hospitalized patients with heart failure aged ≥65 years. The study objectives were to examine the prevalence, overlap, and prognostic implications of the coexistence of multiple frailty domains. Physical frailty, social frailty, and cognitive dysfunction were evaluated by the Fried phenotype model, Makizako's 5 items, and Mini-Cog, respectively. The primary study outcome was the combined endpoint of heart failure rehospitalization and all-cause death within 1 year. Among 1180 enrolled hospitalized patients (median age, 81 years; 57.4% male), physical frailty, social frailty, and cognitive dysfunction were identified in 56.1%, 66.4%, and 37.1% of the patients, respectively. The number of identified frailty domains was 0, 1, 2, and 3 in 13.5%, 31.4%, 36.9%, and 18.2% of the patients, respectively. During follow-up, the combined endpoint occurred in 383 patients. Adjusted hazard ratios for 1, 2, and 3 domains, with 0 domains as the reference, were 1.38 [95% confidence interval (CI) 0.89-2.13; P = 0.15], 1.60 (95% CI 1.04-2.46; P = 0.034), and 2.04 (95% CI 1.28-3.24; P = 0.003), respectively. Incorporating the number of frailty domains into the pre-existing risk model yielded a 22.0% (95% CI 0.087-0.352; P = 0.001) net reclassification improvement for the primary outcome. CONCLUSIONS: The coexistence of multiple frailty domains is prevalent in hospitalized elderly patients with heart failure. Holistic assessment of multi-domain frailty provides additive value to known prognostic factors.