Thirty dogs with experimental myocardial infarction underwent cardiopulmonary bypass, hypothermic asanguineous K+ cardioplegia (1 hour), and reperfusion (30 minutes). Ten hearts were vented throughout, 5 only during arrest, and 5 only during reperfusion; 10 were not vented. Left ventricular (LV) performance and compliance were assessed by isovolumic (LV balloon) indexes before bypass and after reperfusion. Vented hearts recovered 116 +/- 8.3% of prearrest developed LV systolic pressure (DLVSP) and 131 +/- 13.6% of prearrest rate of rise of LV pressure (dP/dt). Nonvented hearts allowed to develop pressure during arrest (11.6 +/- 1.6 mm Hg) and reperfusion (65 +/- 4 mm Hg) recovered 50 +/- 3.9% of prearrest DLVSP and 55 +/- 5% of prearrest dP/dt (p less than 0.05). Reduction in LV compliance was comparable in both groups. Mitochondrial architecture (electron microscopy) was preserved in vented hearts, but was modestly disrupted in nonvented hearts, thus suggesting slight metabolic impairment. Functional recovery was nearly complete in hearts vented only during reperfusion (DLVSP, 94 +/- 10.4%; dP/dt, 89 +/- 12.6%), but venting only during arrest led to functional depression (DLVSP, 50 +/- 6.6%; dP/dt, 51 +/- 8%; p = 0.01). We conclude that venting chronically infarcted hearts during cardiac operations affords better myocardial protection by avoiding the damage that occurs during nonvented reperfusion.