To assess the ability of adenosine technetium-99m sestamibi myocardial perfusion single-photon emission computed tomography (SPECT) to identify high-risk women with severe or extensive coronary artery disease (CAD), we studied 130 consecutive women who underwent adenosine sestamibi myocardial perfusion SPECT and catheterization within 2 months. Severe (> or = 50% stenosis of left main coronary artery, > or = 90% stenosis in the proximal left anterior descending or in > or = 2 coronary arteries) or extensive (> or = 70% stenosis in 3 vessels) CAD was present in 54 patients, whereas 76 had no CAD or mild to moderate CAD. Semiquantitative visual SPECT analysis used 20 segments and a 5-point scoring system (0 = normal, 4 = absent uptake). Among the clinical, hemodynamic and nuclear variables analyzed, univariate predictors of severe or extensive CAD included a higher prescan likelihood of CAD, history of myocardial infarction, a higher heart rate at rest, a lower increase in heart rate during adenosine infusion, a higher summed stress score, summed reversibility score, and multivessel scan abnormality. Multivariate logistic analysis of the most predictive clinical (prescan likelihood of CAD), hemodynamic (increase in heart rate during adenosine infusion), and scan variables (summed stress score) revealed summed stress score (chi-square = 32; p <0.0001) and prescan likelihood of CAD (chi-square = 6.4; p <0.05) as the only independent predictors of severe or extensive CAD. Based on these logistic models, we determined the probability for the presence of severe or extensive CAD in patients with low, intermediate, and high prescan likelihood of CAD across the range of values of a summed stress score. This revealed that there were incremental increases in the probability for severe or extensive CAD both as a function of prescan likelihood of CAD and summed stress score. A severely abnormal scan (summed stress score > 8) during adenosine technetium-99m sestamibi myocardial perfusion SPECT had a high sensitivity of 91% and a moderately high specificity of 70% for identifying high-risk women with severe or extensive CAD. These results coupled with the previously defined prognostic significance of these findings suggest this test to be a useful diagnostic tool for the evaluation of CAD in women.