Although single or multiple bursts of ischemic preconditioning have been shown to protect against myocardial necrosis, the frequency-dependency of the protection afforded by multiple stimuli has not yet been systematically studied in the anesthetized animal. The aim of the present study was to characterize the effect of increasing numbers of preconditioning stimuli (5 min ischemia plus 10 min reperfusion) in a collateral-deficient species in vivo. Rabbits were anesthetized and divided into groups (n = 5-7/group) which were subjected to 0, 1, 2, 4, 6 or 8 cycles of preconditioning (each comprising 5 min of regional ischemia plus 10 min reperfusion) followed by 45 min of regional ischemia and 2 h reperfusion. An additional group, serving as a control, was subjected to eight cycles of preconditioning without a subsequent 45 min of regional ischemia. Infarct size (I) and risk zones (R) were delineated with the aid of tetrazolium staining and fluorescent particles and the I/R ratio was expressed as a percentage. Mean infarct size in the control group which had no preconditioning was 59.8 +/- 4.0%. This was reduced to 22.2 +/- 2.9%, 19.5 +/- 4.1% and 23.3 +/- 3.4% by one, two and four cycles of preconditioning. However, when the number of preconditioning cycles was increased to six or eight protection declined with infarct sizes of 41.8 +/- 6.9% (P < 0.05 v zero preconditioning control group) and 47.1 +/- 7.6% (N.S. v zero preconditioning control group). In the group subjected to eight cycles of preconditioning without subsequent ischemia, infarction was not detected. In conclusion, in the anesthetized rabbit, beyond 4 cycles, protection against infarction diminishes with increasing numbers of preconditioning cycles.