Myocardial scarring and coronary insufficiency give rise to regional changes in left ventricular function, often leading to generalized left ventricular dysfunction during physical activity only. The main purpose of this study was to evaluate the relations between left ventricular function determined by ECG and measurements of pulmonary artery pressure during exercise and the severity of coronary artery disease. Simultaneous measurements in the pulmonary artery and in the left ventricle in 76 patients revealed that the enddiastolic pulmonary pressure (PAEDP) was lower than the enddiastolic pressure in the left ventricle (LVEDP) by a mean value of 10.5 mm Hg in the presence of ventricular dysfunction. This difference was smaller in congestive heart failure than in cases of acute myocardial ischaemia. Correlation coefficients of mean pulmonary wedge pressure (PCm), PAEDP, mean pulmonary pressure (PAPm), and LVEDP were 0.90, 0.86, and 0.81, respectively, thus allowing only an approximate estimate of the left ventricular filling pressure. In 150 angiographically documented cases of coronary heart disease, haemodynamic measurements were performed during stepwise-increased, symptom-limited supine exercise on a bicycle ergometer. All patients limited at 25 watts had either triple vessel disease or stenosis of the trunk of the left coronary artery or of the proximal section of the left anterior descending artery (RIVA). In comparison with subjects with single vessel disease, patients with triple vessel involvement tolerated only a smaller exercise load and reached higher values of PAEDP (30.4 +/- 9.0 versus 24.0 +/- 7.7 mm Hg, p less than 0.001). Analysis of data of patients with a single coronary stenosis showed the exercise-PAEDP to be largely independent of the myocardial condition, but to depend upon the location of the stenosis, the highest pressure values being observed with stenoses of the main left coronary artery or the proximal segment of the RIVA. Based on these findings a simple coronary score system was delineated to determine the severity of the disease, taking into account the location of an obstruction, in particular, and, to a lesser amount, the degree of the stenosis and the type of coronary artery distribution. The score yielded essentially better correlations to work load and filling pressures during exercise than did the number of obstructed vessels. The regression line of the PAEDP versus the coronary score was flatter in patients with angiographically documented collaterals than in cases without, indicating the functional significance of these vessels. In patients with stenoses confined to the arteries supplying infarcted areas and, consequently, without signs of ischaemia during exercise a close relation was obtained between the left ventricular ejection fraction (EF) and the maximum PAEDP, best expressed by a third order regression equation (r = 0.79), p less than 0.001, SEE +/- 6.1). A PAEDP exceeding 25 mm Hg is, thus, a reliable sign of an EF of less than 40%...