To define the clinical and adenosine test variables that predicted perioperative cardiac events, 122 patients who received adenosine radionuclide perfusion imaging before peripheral vascular surgery were reviewed. Events included pulmonary edema, an ischemic end point of acute myocardial infarction (AMI) or cardiac death. Five patients underwent coronary revascularization before the surgical procedure. Of the 117 remaining patients, 19 had pulmonary edema, 10 had an AMI, and 2 died after peripheral vascular surgery. Most of the patients (78%) were in an intermediate-risk group as indicated by the presence of ≥1 clinical risk factor as defined by the Eagle criteria. The only predictor of perioperative pulmonary edema was a history of congestive heart failure (33% vs 4%; p = 0.002). No clinical variables predicted AMI or death. The adenosine variables that were univariate predictors of AMI and death were the number of reversible perfusion defects (1.75 ± 1.84 vs 0.75 ± 0.90; p = 0.001) and the number of coronary artery distributions with a radionuclide perfusion defect (1.33 ± 0.64 vs 0.85 ± 0.67; p = 0.022). The number of reversible perfusion defects was the only multivariate predictor or ischemic events (p = 0.017). the presence or >1 reversible defect was associated with an increased frequency of ischemic events (68% vs 28%; p = 0.045). The sensitivity and specificity of >1 reversible defect was 58% and 73%, respectively, with a positive and negative predictive value of 19% and 94%. We conclude that (1) a history of congestive heart failure is predictive of perioperative pulmonary edema, and (2) the presence or absence of multiple reversible defects on adenosine perfusion images can provide useful information for identifying patients at risk for a perioperative ischemic event.