BACKGROUND: Cardiac preconditioning is thought to be involved in the observed decreased coronary artery reocclusion rate in patients with angina preceding myocardial infarction. We prospectively examined whether preconditioning by sevoflurane would decrease late cardiac events in patients undergoing coronary artery bypass graft (CABG) surgery. METHODS: Seventy-two patients scheduled for elective CABG surgery were randomized to preconditioning by sevoflurane (10 min at 4 vol%) or placebo. For all patients, follow-up of adverse cardiac events was obtained 6 and 12 months after surgery. Transcript levels for platelet-endothelial cell adhesion molecule-1 (PECAM-1/CD31), catalase and heat shock protein 70 (Hsp70) were determined in atrial biopsies after sevoflurane preconditioning. RESULTS: Pharmacological preconditioning by sevoflurane reduced the incidence of late cardiac events during the first year after CABG surgery (sevoflurane 3% vs 17% in the placebo group, log-rank test, P=0.038). One patient in the sevoflurane group and three patients in the placebo group experienced new episodes of congestive heart failure and three additional patients had coronary artery reocclusion. Perioperative peak concentrations for myocardial injury markers were higher in patients with subsequent late cardiac events [NTproBNP, 9031 (4125) vs 3049 (1906) ng litre(-1), P<0.001; cTnT, 1.31 (0.88) vs 0.46 (0.29) microg litre(-1), P<0.001]. Transcript levels were reduced for PECAM-1 and increased for catalase but unchanged for Hsp70 in atrial biopsies after sevoflurane preconditioning. CONCLUSIONS: This prospective randomized clinical study provides evidence of a protective role for pharmacological preconditioning by sevoflurane in late cardiac events in CABG patients, which may be related to favourable transcriptional changes in pro- and antiprotective proteins.