The cardioprotective effects of different aerobic exercises on chronic heart failure with different aetiologies and whether mitophagy is involved remain elusive. In the current research, left anterior descending ligation and transverse aortic constriction surgeries were used to establish mouse models of heart failure, followed by 8 weeks of moderate-intensity continuous training (MICT) and high-intensity interval training (HIIT). The results showed that for ischaemic heart failure MICT significantly improved ejection fraction (P < 0.05) and fractional shortening (P < 0.05), mitigated left ventricular end-systolic dimension (P < 0.01), decreased brain natriuretic peptide (P < 0.0001) and mitigated fibrosis (P < 0.0001), while HIIT only decreased brain natriuretic peptide (P < 0.0001) and fibrosis (P < 0.0001). For pressure-overload heart failure, both MICT and HIIT significantly increased ejection fraction (P < 0.0001) and fractional shortening (MICT: P < 0.001, HIIT: P < 0.0001), and reduced left ventricular end-diastolic and end-systolic dimensions, brain natriuretic peptide (P < 0.0001), and fibrosis (MICT: P < 0.01, HIIT: P < 0.0001); HIIT was even better in reducing brain natriuretic peptide. Myocardial autophagy and mitophagy were compromised in heart failure, and the exercises improved myocardial autophagic flux and mitophagy inconsistently in heart failure with different aetiologies. Significant correlations were found between multiple mitophagy pathways and the cardioprotection of the exercises. Collectively, MICT may be the 'optimum' modality for ischaemic heart failure, while both MICT and HIIT (especially HIIT) were suitable for pressure-overload heart failure. Exercises differently improved myocardial autophagy/mitophagy, and multiple mitophagy-related pathways were closely implicated in cardioprotection of exercises for chronic heart failure.