The effects of two types of repeated transient coronary artery occlusions on collateral development were examined in chronically instrumented, conscious dogs. A 2-minute coronary occlusion at 32-minute intervals (group 1, n = 11) or a 15-second occlusion at 4-minute intervals (group 2, n = 7) were repeated day and night without interruption. In both groups, the total duration of coronary occlusions each day was the same (90 minutes). Before and after repetitive occlusions of either group, effects of transient 2-minute coronary occlusion on regional segment shortening in the ischemic area were examined to assess the functional state of the collateral vessels. In group 1, systolic segment shortening in the area rendered ischemic was reduced to -97.8 +/- 17.7% of the preocclusive control value during 2 minutes of coronary occlusion. After 125-478 repetitive occlusions (3-11 days), the degree of hypokinesia during the 2-minute occlusion was significantly improved to -0.6 +/- 4.6% of the preocclusive value (p less than 0.001 vs. before the repetition). In group 2, it remained unchanged even after 3,500-5,450 repetitive occlusions (11-16 days): -111.8 +/- 8.2% before and -111.4 +/- 13.8% after the repetition of 15-second occlusions (NS). The ratio of peripheral coronary arterial pressure to aortic pressure during transient-coronary occlusion, measured by selective catheterization, was significantly higher in group 1 than in group 2 (64.4 +/- 5.3% vs. 20.7 +/- 1.3%, p less than 0.001). These findings suggest that myocardial ischemia of 2 minutes but not 15 seconds is vital to provide effective stimuli for angiogenesis.