Regional and overall left ventricular contraction reserve was studied in 14 patients with coronary heart disease, in 5 healthy subjects and in 4 patients before and after aorto-coronary bypass surgery. Quantification of overall contraction was based on ventricular volumes and ejection fraction. Regional contraction reserve was calculated with the hemiaxis method and a ventricular score. Contraction reserve under nitroglycerin and in postextrasystolic beats was compared. For routine quantification of contraction reserve the ventricular score is recommended. For research purposes the hemiaxis method is to be preferred. Postextrasystolic beats are better suited for analysis of contraction reserve than are angiograms following administration of nitroglycerin. This is due to the minor expense of the procedure, furthermore, postextrasystolic beats allow better differentiation between contracting and non-contracting areas. Left ventricular contraction reserve is larger in patients with coronary heart disease, angina pectoris and ischemic reactions in the exercise ECG than in control patients. These findings are based on overall and on regional volume parameters. A quantitatively greater improvement in contraction could be provoked in the anterior wall than in the posterior wall. Regional contraction improved significantly in most cases either in the anterior wall or in the posterior wall; rarely it improved simultaneously in both left ventricular regions. In a few cases contraction deteriorated in one area with a simultaneous improvement in the opposite area. Overall and regional ventricular function, as assessed preoperatively by contraction reserve determinations could not be completely regained in normal beats after successful bypass surgery. Differences in the regional contraction reserve seemed to be mainly due to varying degrees of ischemia and scarring.