Background: Physical activity (PA) is an important component of secondary prevention after myocardial infarction (MI). The mortality risk of MI survivors varies at different post-MI periods, yet the time-varying effect of total PA is unclear. We aimed to investigate the association between different volumes and patterns of total PA and mortality at different post-MI periods. Methods: Using data from the China Patient-centered Evaluative Assessment of Cardiac Events Million Persons Project, we divided the screened MI survivors into within-1-year and beyond-1-year groups based on the duration between their baseline interview and MI onset. Total PA was divided into insufficient (3000 metabolic equivalent of task [MET] minutes/week) and sufficient PA. Sufficient PA was further categorized as moderate and high (3000–4500 and >4500 MET minutes/week) volumes; leisure (\ge50%) and non-leisure (>50%) patterns. Data on mortality were derived from the National Mortality Surveillance System and Vital Registration of the Chinese Center for Disease Control and Prevention. Cox proportional hazard models were fitted to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). Restricted cubic spline regression analyses were performed to examine the dose-response association between PA and mortality. Results: During the follow-up (median 3.7 years) of the 20,653 post-MI patients, 751 patients died. In the within-1-year group, moderate (HR: 0.59, 95% CI: 0.40 to 0.88) and high (0.63, 0.45 to 0.88) volumes and both patterns (leisure: 0.52, 0.29 to 0.94; non-leisure: 0.64, 0.46 to 0.88) of PA were all associated with significantly lower risk of mortality, compared with insufficient PA. In the beyond-1-year group, the association was observed in high volume (0.69, 0.56 to 0.86) and both patterns (leisure: 0.64, 0.48 to 0.87; non-leisure: 0.79, 0.65 to 0.97). A non-linear relationship between PA and mortality was found in the within-1-year group (p for non-linearity 0.001), while a linear relationship was demonstrated in the beyond-1-year group (p for non-linearity = 0.107). Conclusions: Sufficient total PA was associated with mortality risk reduction after MI, either leisure or non-leisure pattern. Different dose-response associations between PA and mortality were found at different post-MI periods. These results could promote individualized and scientifically derived secondary prevention strategies for MI.