AIMS: Frailty in older patients with stage B heart failure with preserved ejection fraction (HFpEF) has not been fully explored. We evaluated the prevalence and prognostic significance of frailty in older patients diagnosed with stage B HFpEF. METHODS: Our prospective cohort study included inpatients aged ≥65 years who were followed up for 3 years. Stage B HFpEF was defined as cardiac structural or functional abnormalities with a left ventricular ejection fraction (LVEF) ≥ 50% without signs or symptoms. Frailty was assessed using the Fried phenotype. The primary outcome was 3-year all-cause mortality or readmission. RESULTS: Overall, 520 older inpatients diagnosed with stage B HFpEF [mean ± standard deviation age: 75.5 ± 6.25 years, male: 222 (42.7%)] were included in the study. Of these, 145 (27.9%) were frail. Frail patients were older (78.5 ± 6.23 vs. 74.3 ± 6.22 years, P < 0.001), with a lower body mass index (24.6 ± 3.60 vs. 25.7 ± 3.27 kg/m , P = 0.001), higher level of N-terminal pro-B-type natriuretic peptide [279 (interquartile range: 112.4, 596) vs. 140 (67.1, 266) pg/mL, P < 0.001], longer timed up-and-go test result (19.9 ± 9.71 vs. 13.3 ± 5.08 s, P < 0.001), and poorer performance in the short physical performance battery (4.1 ± 3.26 vs. 8.2 ± 2.62, P < 0.001), basic activities of daily living (BADL, 4.7 ± 1.71 vs. 5.7 ± 0.57, P < 0.001), and instrumental activities of daily living (IADL, 4.4 ± 2.73 vs. 7.4 ± 1.33, P < 0.001). Frail patients were more likely to have a Mini-Mental State Examination (MMSE) score <24 (55.9% vs. 28.8%, P < 0.001) and take more than five medications (64.1% vs. 47.2%, P = 0.001). Frail patients had a higher incidence of all-cause mortality or readmission (62.8% vs. 47.7%, P = 0.002), all-cause readmission (56.6% vs. 45.9%, P = 0.029), and readmission for non-heart failure (55.2% vs. 41.3%, P = 0.004) during the 3-year follow-up, with a 1.53-fold (95%CI 1.11-2.11, P = 0.009) higher risk of all-cause mortality or readmission, a 1.52-fold (95%CI 1.09-2.11, P = 0.014) higher risk of all-cause readmission, and a 1.70-fold (95%CI 1.21-2.38, P = 0.002) higher risk of readmission for non-clinical heart failure, adjusted for sex, age, polypharmacy, Athens Insomnia Scale, MMSE, LVEF, BADL, and IADL. CONCLUSIONS: Frailty is common in elderly patients with stage B HFpEF. Physical frailty, particularly low physical activity, can independently predict the long-term prognosis in these patients.