BACKGROUND: Reduced septal or anteroseptal uptake of thallium-201 during exercise is frequently observed in patients with left bundle branch block (LBBB) even in the absence of left anterior descending (LAD) coronary artery disease. The purpose of this study was to evaluate prospectively the accuracy of dipyridamole 201Tl single-photon emission computed tomography (SPECT) in detecting LAD coronary artery disease in patients with LBBB and septal or anteroseptal perfusion defects on exercise 201Tl SPECT. METHODS AND RESULTS: Twelve consecutive patients (10 men and two women) with complete LBBB and septal or anteroseptal perfusion defects on exercise 201Tl SPECT underwent dipyridamole 201Tl SPECT. The delay between dipyridamole and exercise was 2 to 30 days. Coronary angiography was performed during this period in all patients. Six (50%) of 12 patients with exercise perfusion defects showed normal perfusion after dipyridamole; all had normal coronary angiograms. The remaining six patients also had positive results of dipyridamole studies, two with moderate and four with severe septal or anteroseptal perfusion defects. Coronary angiography showed significant (> 50%) LAD coronary artery stenosis in three patients; three patients with severe septal or anteroseptal perfusion defects after dipyridamole had normal coronary angiograms. Neither the evaluation of apical involvement nor the presence of dilated ventricles, decreased left ventricular ejection fraction, or wall motion abnormalities could help to identify (or explain) false-positive results. CONCLUSION: This study confirms that dipyridamole is more accurate than exercise in excluding LAD coronary artery disease. However, there are still false-positive results and the severity of the septal or anteroseptal perfusion defect does not add additional information to identify LAD coronary artery disease. Coronary angiography is thus necessary for positive dipyridamole study results to identify coronary artery disease as a major prognostic factor in patients with LBBB.