Dipyridamole thallium-201 myocardial imaging can provide information regarding risk of perioperative cardiac events in patients being considered for vascular surgery. The value for this purpose of myocardial imaging with technetium-99m sestamibi (MIBI), a radiotracer with biokinetic and imaging properties different from thallium-201, has not been established. To this end the prognostic value of dipyridamole MIBI tomography for perioperative and late cardiac events was evaluated in 229 consecutive patients being considered for elective vascular surgery. Vascular surgery was done < or = 3 months after testing in 197 of these patients. Perioperative cardiac events (cardiac death, nonfatal myocardial infarction, unstable angina, or ischemic pulmonary edema) occurred in 9 (5%) patients. The rate of such events was 3% in patients with normal MIBI results, 5% in those with abnormal results, and 6% in patients with a reversible MIBI defect (both p = NS). When patients with abnormal MIBI results who had preoperative cardiac interventions (coronary revascularization or an increase in antiischemic medical therapy) were compared with with those who did not, no significant differences in the occurrence of perioperative cardiac events were found between these two groups either. A group of 172 medically treated patients who survived vascular surgery and did not have a nonfatal perioperative cardiac event was then monitored (mean 21 +/- 14 months) for the occurrence of a serious late cardiac event (nonfatal myocardial infarction or cardiac death). Event-free survival (Mantel-Cox) was significantly less in patients with abnormal studies compared with those with normal scan results. Late cardiac events occurred in 26 (15%) patients, with those having an abnormal MIBI result showing a significantly greater event rate than those with normal results (26% vs 4%, p < 0.0001). The rate of late cardiac events was 33% in patients with a reversible MIBI defect (p < 0.001) and 23% in those with a fixed defect (p < 0.03). Independent Cox multivariable predictors of increased risk of late cardiac events were a history of diabetes mellitus (relative risk [RR] 2.2, 95% confidence interval [CI] 1.0 to 4.9), an abnormal MIBI study (RR 3.7, 95% CI 1.2 to 11.4), and a reversible MIBI defect (RR 2.7, 95% CI 1.2 to 6.1). We conclude that, although its ability to assess increased perioperative cardiac risk remains uncertain, dipyridamole MIBI tomography does provide important prognostic information regarding the risk of serious cardiac events in patients having vascular surgery. The presence of an abnormal MIBI result, specifically one demonstrating a reversible perfusion defect, is associated with significantly increased risk.