Background While physical activity reduces risk of developing myocardial infarction (MI), it is unknown whether a history of physical activity is also protective of fatal arrhythmia and case-fatality in patients who have suffered an acute MI. Methods 104,801 individuals included in 2003–2014 in the Copenhagen General Population Study (CGPS), a prospective population-based study with self-reported leisure time physical activity (LTPA) in three categories measured at baseline, were followed until 2014 through national registries. The 1,517 individuals who suffered a first time MI during follow-up constituted the study population. Outcomes were fatal MI, defined as date of death same as date of MI (including out-of-hospital deaths) and 28-day fatality. Through multivariable analyses the association between baseline LTPA and outcomes were assessed adjusted for CVD risk factors. Results Of 1,517 MI events, 117 (7.7%) were fatal and another 79 (5.6%) lead to death within 28 days. Median time from baseline to MI was 3.6 years (IQR 1.7–5.8). LTPA was associated with lower risk of fatal MI with odds ratios of 0.40 (95% CI: 0.22–0.73) for light and 0.41 (0.22–0.76) for moderate/high LTPA after multivariable adjustment with sedentary LTPA as reference. Age, alcohol-intake, education and smoking were identified as other predictors for fatal MI. We found no association between LTPA and 28-day case fatality. Conclusions Among individuals with MI, those that have engaged in any light or moderate physical activity were more likely to survive their MI. Results are consistent with effect of exercise preconditioning on risk of fatal arrhythmia.