The effects of digitalis (ouabain infusion, priming dose of 10 micrograms/kg/min followed by 2.0 micrograms/kg/min) on subepicardial and subendocardial ischemic dysfunction during partial occlusion (59.4% reduction in coronary flow) and total coronary occlusion were evaluated in 14 dogs using pairs of ultrasonic crystals implanted in the epicardial and endocardial layers. After 30 minutes of partial coronary occlusion, systolic shortening in the ischemic zone decreased from 12.4 +/- 2.9% to -0.4 +/- 0.9% (p less than 0.001) in the epicardium and from 18.9 +/- 3.1% to -1.0 +/- 1.1% (p less than 0.001) in the endocardium. Ouabain infusion increased the systolic shortening from -0.4 +/- 0.9% to 10.6 +/- 3.1% (p less than 0.001) in the epicardium and from -1.0 +/- 1.1% to 17.2 +/- 3.4% (p less than 0.001) in the endocardium. The end-diastolic length did not change. In contrast, after 30 minutes of total coronary occlusion, systolic shortening in both epicardial and endocardial layers was replaced by systolic lengthening and remained unaffected by ouabain infusion in both layers. Systolic shortening in the nonischemic epicardial and endocardial layers increased consistently. We conclude that ouabain improves the contractile function of both ischemic epicardial and endocardial layers layers after partial coronary occlusion and does not worsen subendocardial ischemic dysfunction. After total coronary occlusion, however, the contraction of the ischemic zone is unaffected by ouabain.