Objectives: Induced tolerance to bacterial lipopolysaccharide (LPS) by pretreatment with sublethal doses of LPS has been shown to reduce the inflammatory response of monocytes, circulating PMNs and PMN adhesion to endothelial cells in response to subsequent stimuli, and also to increase cellular and organ tolerance to stress by other mechanisms. Therefore, we undertook to determine whether or not LPS desensitization is associated with reduced myocardial infarct size at 3 days after reperfusion following coronary occlusion. Methods: Rats were randomized to either daily intraperitoneal LPS injections to provide LPS tolerance, or to equal volumes of saline (controls). In both groups at day 7 nontransmural infarction was produced by a 45 min coronary occlusion followed by 3 days of reperfusion during which LPS injections were continued. Histologic infarct size was assessed as percent of the left ventricle and as a percent of the risk zone (determined by fluorescent microspheres). Results: Myocardial infarct size as percent of the left ventricle and of the risk zone were significantly reduced in the LPS-tolerant group (n = 14) compared to control rats (n = 12), the latter being reduced by 37% (33.6 +/- 18.4 vs. 54.1 +/- 8.6% of the risk zone, P < 0.002). The percentages of activated circulating PMN after LPS desensitization and saline pretreatment were not different prior to coronary occlusion (at 7 days), but 3 days after coronary occlusion and reperfusion the percent of activated PMNs in the treated group was markedly reduced compared to controls (2.9 +/- 1.6 vs. 11.4 +/- 7.2%, respectively, P < 0.02). Conclusions: LPS desensitization in rats for 1 week prior to coronary occlusion inhibited activation of circulating PMNs 3 days after reperfusion following 45 min of coronary occlusion. LPS also is well-known to induce heat stress proteins and may affect other protective mechanisms. These actions are associated with a significant reduction in myocardial infarct size in LPS-tolerant animals compared to untreated controls.