Exercise-induced increases in pulmonary uptake of thallium-201 (201Tl) have been associated with exercise-induced myocardial dysfunction. To evaluate this phenomenon more replicably, a quantitative semi-automated computer program was used to generate, from anterior exercise and delayed views, lung-myocardial ratios (LMR) of201Tl uptake in 78 patients [40 normal, 38 with coronary artery disease (CAD)]. Patients with CAD had a significantly higher mean exercise lung myocardial ratio (EXLMR) than normals (30.8 vs. 27.3;P < 0.003). In patients with adequate exercise (≥85% of an age-adjusted maximal heart rate), the EXLMRs of CAD patients were significantly higher than those of normals (29.7 vs. 25.5;P=0.003). However, this difference between CAD and normal patients was not apparent in a patient subgroup with submaximal exercise levels (< 85% of an age-adjusted maximal heart rate). In both normal and CAD patients, EXLMR decreased with increasing exercise levels (r=-0.555;P=0.007). In patients with201Tl scans lacking visually defined perfusion defects (visually normal), an elevated LMR detected 60% of CAD cases with 81% specificity. A considerably elevated EXLMR in patients achieving adequate exercise should suggest the presence of CAD, even if there are no visually apparent cardiac perfusion defects. With submaximal exercise, however, the EXLMR is not a useful discriminator between CAD patients and normals.