SUMMARY Based on a prospective study in 200 consecutive patients with myocardial infarction, we reported previously that early recurrent myocardial infarction is more frequent after nontransmural than transmural infarction. Multiple logistic regression analysis using 14 clinical variables identified, in addition to type of infarction (nontransmural), three other risk factors for early recurrent infarction: obesity, female gender and recurrent chest pain. Early recurrent infarction was documented by reelevation in plasma MB CK activity. The present study was performed to assess the accuracy of these variables as prospective predictors of early recurrent infarction. Studies were performed in a new population of 150 patients admitted consecutively with acute myocardial infarction (test set). The regression coefficients derived by multiple logistic analysis from the training set population were applied by comparable analysis to the test set population and the presence or absence of recurrent infarction was predicted correctly for 80% of the patients. Patients were followed prospectively for 9 months (range 3-18 months) and life-table analysis was performed to assess the impact of recurrent infarction on short and long term survival. During the first 21 days after infarction, mortality was 23% for patients with transmural and 10% for nontransmural infarction (p < 0.01). However, among patients with nontransmural infarction, hospital mortality was 23% among those with early recurrent infarction and 8% for those without recurrence (n < 0.05). At the conclusion of follow-up, mortality was 34% among patients with nontransmural infarction and recurrence, compared with 23% among patients with nontransmural infarction and no recurrence. Thus, early recurrent infarction has a distinct deleterious effect on survival, and a subset of patients more likely to experience recurrent infarction can be prospectively identified.EARLY recurrent myocardial infarction, or extension of infarction, has been recognized clinically for many years as manifested by severe prolonged recurrent chest pain and new ST-T-wave abnormalities on the ECG.' Results of earlier studies suggested an incidence of early recurrent infarction of up to 86% ,2 but recent studies using objective specific enzymatic criteria have reported an incidence of 17-25%).6 In a previous report of 200 consecutive patients with acute myocardial infarction, early recurrent infarction detected by reelevation of plasma MB creatine kinase (MB CK) was observed in 17% of patients with documented myocardial infarction.7 8Analyses were performed to determine whether certain clinical characteristics of the population might identify a subset of patients at high risk for subsequent recurrent infarction. Those who experienced an initial nontransmural myocardial infarction comprised one subset of patients at high risk. high risk of developing early recurrent infarction. Obesity was defined as at least 25% in excess of ideal body weight for the person's height and recurrent pain was defined ...