The purpose of this clinical study was to (1) evaluate mortality rates after surgical interventions for patients in cardiogenic shock (CS) secondary to acute coronary occlusion, acute ventricular septal defect (VSD) or acute valvular heart disease, (2) determine the pre-operative regional wall motion, and (3) ascertain the recovery of preoperative regional wall motion abnormalities after surgical intervention. The hospital records of twenty-five consecutive patients in CS were reviewed retrospectively. Regional wall motion was assessed preoperatively by ventriculography and postoperatively by 2D echocardiography (Sonotron Kardio VUE 60) after 1 and 3 days and at the day of discharge from the surgical ward (7-10 days). The left ventricle was divided in three segments according to the blood supply: LAD artery (antero-lateral wall), circumflex artery (lateral wall), and right coronary artery (inferior and basal wall). Regional wall motion was analyzed with the use of a scoring system in which grading was from 0 to 4 according to the following criteria: 0 = hyperkinesia, 1 = normokinesia, 2 = hypokinesia, 3 = dyskinesia, 4 = akinesia. Postmortem examination was performed in 8/9 patients. Data are presented as mean +/- SD. Significant differences were defined as probabilities for each test of p less than 0.05. The hospital mortality was higher for patients with acute coronary occlusion as compared to those with acute valvular disease or VSD (54.5%, 27.3%, 0%, resp.). The cause of death was cardiac in 7/9 patients. However, postmortem examination revealed loss by infarction of only moderate quantities of myocardium which could not explain the severe postoperative heart failure in those patients. Previous myocardial infarctions and preoperative cardiac arrest were significant risk factors for hospital mortality. In all patients with acute coronary occlusion (11/11) at least one region of the left ventricle was either a- or dyskinetic in the region supplied by the acute occluded vessel. In addition five patients had akinetic regions due to previous infarctions. The remaining remote myocardium was hypocontractile due to significant stenosis in coronary arteries supplying remote areas. Of 10 dyskinetic segments before surgical intervention, 5 were akinetic postoperatively, and only 5 developed slight hypokinetic contractions. The overall hypokinetic regions were not different as compared to the preoperative data (36.4% vs 39.4%). The normokinetic segments increased from 9.1% to 33.% (p less than 0.05).