The incidence, outcome and predictors of the in-hospital development of cardiogenic shock and its prognostic significance were analyzed in 845 patients presenting with acute myocardial infarction. Cardiogenic shock developed after hospitalization in 60 patients (7.1%). In half of these patients, cardiogenic shock developed at least 24 h after hospital admission. The in-hospital mortality rate was greater than 15 times higher for patients with cardiogenic shock than for patients without shock (65.0% versus 4.3%, respectively, p less than 0.001). Enzymatic evidence of infarct extension occurred in 23.3% of the patients with shock compared with 7.4% of those without shock (p less than 0.0001). Multivariate analysis indicated that independent predictors for the in-hospital development of cardiogenic shock were age greater than 65 years (p = 0.007), left ventricular ejection fraction on hospital admission less than 35% (p = 0.007), large infarct as estimated from serial enzyme determinations (that is, peak creatine kinase-MB isoenzyme greater than 160 IU/liter (p = 0.008), history of diabetes mellitus (p = 0.011) and previous myocardial infarction (p = 0.012). Patients with three, four or five of these risk factors had a 17.9%, 33.7% or 54.4% probability, respectively, of developing cardiogenic shock after hospital admission. Left ventricular function, as reflected by left ventricular ejection fraction (p = 0.04) and severity of left ventricular wall motion abnormality (p = 0.04), was the most important determinant of in-hospital mortality in the patients with cardiogenic shock.