Myocardial perfusion imaging with thallium is an established method to assess the presence of ischaemic and viable myocardium. Using planar imaging, images acquired 1 h after immediate thallium re-injection have been shown to be comparable to those of 3 h redistribution. The aim of this study was to clarify this using SPET. Twenty patients with chronic coronary artery disease with at least two perfusion defects on post-exercise images participated in this study. All patients received thallium re-injection after post-stress SPET and all had 1 h, 3 h and 24 h redistribution imaging. The thallium myocardial tomograms were divided into 14 segments for each patient. A total of 78 segments were studied. When the frequency of reversibility on 1 h, 3 h and 24 h redistribution images is compared, of the 78 SPET defects 18 (23.1%) showed reversibility by 1 h, 34 (43.6%) by combined 1 h and 3 h imaging (P < 0.05) and 50 (64.1%) by combined 1 h, 3 h and 24 h imaging (P < 0.05). Our results show that, compared to 3 h images, images acquired 1 h after immediate thallium re-injection underestimate the extent of viable myocardial segments and the incidence of late reversibility was not reduced by the immediate re-injection of thallium.