Objective：The number of heart failure （HF） patients is increasing in Japan as its population continues to age, but little is known about current medication strategies. We investigated the relationship between medication changes during hospitalization and the readmission rate among older Japanese patients with new-onset HF.Design：Retrospective cohort study.Methods：We analyzed medical record data from Toho University Medical Center Omori Hospital between March 2004 and April 2018. Initial admissions for new-onset HF in patients aged≥75 years were examined (n＝329). The class Ⅰ recommended medications stipulated in the JCS 2017/JHFS 2017 guidelines were used as the target medications for this study. Patients with dose titrations or additions of the target medications during hospitalization (dose titrations or additions group) were compared with patients without these changes (the other group). The primary outcome was readmission due to HF within one year of discharge. A hazard ratio, adjusted for potential confounders, was estimated using a Cox proportional hazards model.Results：There were 231 patients in dose titrations or additions group and 98 patients in the other group.The one-year readmission rate was 26.5％ in dose titrations or additions group and 31.8％ in the other group. The adjusted hazard ratio of medication changes for readmission was 0.82 (95％ confidence interval, 0.51-1.33, P＝0.415), but was not statistically significant.Conclusion：The older HF patients in dose titrations or additions group showed a reduced risk of readmission, but lacked significance due to low statistical power.