Background: We identified variation in delivery of guideline recommended care at our institution, and undertook a project to design a heart failure (HF) model of care., Aim: To maximise time patients with HF spend well in the community by delivering best practice guidelines to reduce variation in care improving overall outcomes., Methods: This quality improvement project focused on reducing variation in process measures of care. The HF model of care included electronic HF care bundles, a patient education pack with staff training on delivering HF patient education, referral of all HF patients to the Hospital Admissions Risk Program for phone call within 72 h, and a nurse-pharmacist early follow-up clinic. Outcomes were assessed using interrupted time series analyses., Results: The pre-intervention group comprised 1585 patients, and post-intervention 1720 patients with a primary diagnosis of HF admitted under general cardiology and general medicine. Interrupted time series analysis indicated 30-day readmissions did not change in overall trend (-0.2% per month, P = 0.479) but a significant immediate step-down of 7.8% was seen (P = 0.018). For 90-day readmissions, a significant trend reduction over the time period was seen (-0.6% per month, P = 0.017) with a significant immediate step-down (-9.4%, P = 0.001). Emergency department representations, in-patient mortality and length of stay did not change significantly. Improvements in process measures were seen at audit., Conclusion: This model of care resulted in overall trends of reductions in 30- and 90-day readmissions, without increasing emergency department representations, mortality and length of stay. This model will be adapted as the electronic medical record is introduced at our institution., (© 2020 Royal Australasian College of Physicians.)