This study was designed to assess the contribution of clinical, electrocardiographic and exercise hemodynamic variables to the prediction of normalization on resting reinjection scintigraphy of persistent thallium-201 (<sup>201</sup>Tl) myocardial perfusion defects seen with exercise and 2- to 4-hour delayed (redistribution) imaging. To evaluate this contribution, we studied 159 consecutive patients with persistent <sup>201</sup>Tl myocardial perfusion defects on routine exercise and 2- to 4-hour-delayed scintigrams at the University of Rochester Medical Center who were classified as having moderate or greater ischemic normalization (group 1, n = 76) or minimal to no ischemic normalization (group 2, n = 83) by reinjection scintigraphy. Multiple logistic regression analysis with backward elimination was used to model the effects of clinical, electrocardiographic and exercise hemodynamic data on the odds ratio of a normalized defect. No difference was observed in the two groups with regard to gender, angina on exertion, rate-pressure product, exercise duration, resting or exertional ischemic ST changes on electrocardiogram, presence of Q waves or left ventricular hypertrophy on baseline electrocardiogram, or total number of stress thallium defects (2.8 ± 1.5 segments). No single variable or combination of variables discriminated between groups 1 and 2 by logistic regression analysis. We conclude that defect normalization seen on resting <sup>201</sup>Tl myocardial perfusion scintigraphy is prevalent in patients with persistent defects on routine exercise and delayed myocardial perfusion scintigraphy, and was not predictable from available clinical, electrocardiographic and exercise hemodynamic variables.