Readmissions within 30 days after acute myocardial infarction have been used as a performance metric for hospitals. However, evolving concepts of value-based reimbursement have shifted the focus to 90 days after hospital discharge. Tools are needed to determine risk for 90-day readmission to identify patients who might benefit from enhanced transitional healthcare resources. In this cohort study, we identified all Medicare beneficiaries with a primary diagnosis of acute myocardial infarction who were discharged from hospitals participating in National Cardiovascular Data Registry ACTION registry between 2008 and 2014. Among a random 70% sample (derivation cohort), we performed hierarchical proportional hazards regression, accounting for death as a competing risk, to assess predictors of all-cause readmission within 90 days. Models were validated in the remaining 30%. Among 86 849 unique patients, 23 912 (27.5%) were readmitted within 90 days. Of the readmissions, 55% occurred within 30 days and 81% occurred within 60 days. Predictors of readmission included older age and a history of diabetes mellitus or heart failure. Coronary revascularization was associated with a lower risk of readmission. A simple risk score incorporating patient demographic and clinical characteristics known before discharge identified groups of patients with readmission risks ranging from 13.1% to 42.9%. Model discrimination was moderate (C statistic=0.662), and calibration was excellent (slope=0.97, intercept=-0.04). Readmission within 90 days of hospitalization for acute myocardial infarction can be predicted by variables known before discharge and offers the potential to prospectively design transitional care to the risks of individual patients.