Objective To explore the effects of home-based exercise rehabilitation on cardiac structure, valvular function, and exercise capacity in patients with severe aortic stenosis (AS) after transcatheter aortic valve replacement (TAVR). Methods 49 patients with severe AS who underwent TAVR at Zhongshan Hospital, Fudan University, from January 2024 to February 2025 were enrolled. They were divided into an exercise group (n＝25) or a non-exercise group (n＝24) based on participating or not in home-based rehabilitation after TAVR. The exercise group received 12 weeks of home-based exercise training (aerobic exercise plus resistance training every week); the non-exercise group received routine care. Transthoracic echocardiography (TTE) was used to assess cardiac structural parameters before discharge (T0) and after 12 weeks of exercise (T1). Functional outcomes including the 6-minute walk test (6MWT), Duke Activity Status Index (DASI), and Short Physical Performance Battery (SPPB) were compared between the two groups. A linear mixed-effects model was used to analyze the effect of home-based rehabilitation on echocardiographic parameters. Patients were stratified by baseline 6MWT (＜240 m as low-function subgroup, ≥240 m as high-function subgroup) to compare exercise-related outcomes between subgroups. Results At T1, the exercise group had a longer 6MWT distance than the non-exercise group (P＝0.012). The linear mixed-effects model showed that after 12 weeks of exercise, the left ventricular end-diastolic diameter (LVEDD) decreased in the exercise group but slightly increased in the non-exercise group, with a significant difference in changes over time between the two groups (P_interaction＝0.030). The exercise group also showed greater improvement in effective orifice area index (P_interaction＝0.028) and effective orifice area (P_interaction＝0.042) than the non-exercise group. Subgroup analysis revealed that in the low-function subgroup, the exercise group showed greater improvement in the 6MWT (P_interaction＝0.035) and the effective orifice area index (P_interaction＝0.046) compared to the non-exercise group; in the high-function subgroup, the exercise group showed greater improvement only in LVEDD compared to the non-exercise group (P_interaction＝0.046). Conclusions Home-based exercise rehabilitation improves exercise capacity, optimizes left ventricular remodeling, and enhances valvular function in patients with severe AS after TAVR, with greater benefits observed in patients with lower baseline 6MWT.