Objective: To compare animal models prepared by three different protocols and to explore a stable, reliable and reproducible mouse model of chronic heart failure. Methods: Twenty-five male C57BL/6J mice were randomly divided into four groups, namely blank group (ZC), model A group (MA), B group (MB) and C group (MC). The model group adopted different preparation protocols for continuous injection of ISO, in which the MA group and the MB group were dose-decreasing modeling, and the MA group (10 mg·kg-1 on the first day, 5 mg·kg-1 on the second day, 2.5 mg·kg-1·d-1 on the 3rd~30th day; subcutaneous injection for 30 days); MB group (20 mg·kg-1 on day 1, 10 mg·kg-1 on day 2, 5 mg·kg-1·d-1 on day 3~14; Subcutaneous injection for 14 days), MC group (constant dose 7.5 mg·kg-1·d-1, intraperitoneal injection for 28 days), CHF model was constructed. The day after the end of injection, the survival and mold formation rates of each group of mice were calculated. Cardiac function was measured by cardiac ultrasound and serum levels of N-Terminal Pro-Brain Natriuretic Peptide (NT-pro BNP), interleukin 6 (IL-6), and tumor necrosis factor (TNF-α) were measured in serum. Results: After the end of injection at the end of the fourth week, all model groups were effective in inducing CHF. However, based on the comprehensive test results, it was found that the mold-making situation of the MC group with a concentration of 7.5mg/kg was the most stable, which was more suitable for subsequent research on traditional Chinese medicine. Conclusion: ISO prepared mouse CHF model with constant 7.5 mg·kg-1·d-1 and continuous intraperitoneal injection for 28 days.