A controlled study was undertaken to quantitate and compare the effect of left ventricular bypass (LVB) and left atrial bypass (LAB) on left ventricular infarct volume (LVIV). After baseline studies, the left anterior descending coronary artery in each of 30 mongrel dogs was ligated 1-1.5 cm from its origin. After baseline ischemic studies, control dogs (group 1 10 dogs), LAB dogs (group 2 10 dogs), and LVB dogs (group 3 - 10 dogs) were monitored for four hours. Final infarct size was determined by the nitroblue tetrazolium staining technique. Heart rate, mean arterial pressure, and total systemic flow (TSF) showed no significant difference between control and left heart bypass groups. In group 1, the LVIV was 27.7 ± 6.5 g/ 100 g left ventricle (LV). In group 2, left heart bypass (LHB) flow was 90 ± 4% of TSF. The pressure time index (PTI) was 2845 ± 52 mm Hg-sec/min. The PTI demonstrated no significant difference from controls. In group 2, LVIV was 22.5 6.0 g/100 g LV. LVIV was reduced 18.8% from controls (p < 0.08). In group 2, LHB was complete. Left ventricular decompression (group 3) resulted in a PTI of 328 ± 76 mm Hg-sec/min. The PTI was significantly different (p < 0.001) from groups 1 and 2. The LVIV was 12.6 ± 5.1 g/100 g LV. LVIV was reduced 54.5% from controls (p < 0.001) and 44.0% from group 2 (p < 0.001). These results suggest that LVB may be useful, not only in supporting the circulation in the patient with myocardial infarct and cardiogenic shock, but also in limiting infarct size.