The evaluation of technetium-99m methoxy isobutylisonitrile for the diagnosis of coronary artery disease requires comparative validation against thallium-201, the established perfusion imaging agent. We have compared myocardial and lung uptake of both radiotracers following maximal exercise in 52 patients: 40 with angiographically proven coronary disease. Qualitative and quantitative image analysis showed the diagnostic sensitivity of technetium-99m methoxy isobutylisonitrile to compare favourably with that of thallium-201 as reflected by the mean number of ischaemic segments identified: 5.6 +/- 2.5 vs 4.8 +/- 2.1 by qualitative analysis, and 5.7 +/- 3.2 vs 5.0 +/- 2.6 segments by quantitative analysis. More reversibly ischaemic segments per patient were identified with technetium-99m methoxy isobutylisonitrile than with thallium-201: 3.6 +/- 2.3 vs 1.8 +/- 1.9. There was a higher exercise myocardial to background count ratio with technetium-99m methoxy isobutylisonitrile: 3.16:1 vs 2.58:1, and the mean exercise lung uptake normalised to left ventricular uptake ('lung index'), was lower for technetium-99m methoxy isobutylisonitrile than for thallium-201 (36 +/- 8% vs 40 +/- 10%). Five of the six patients with abnormal elevation of the thallium-201 exercise lung index also had elevation of the technetium-99m methoxy isobutylisonitrile exercise lung index, and all had extensive coronary artery disease. These results indicate that technetium-99m methoxy isobutylisonitrile is at least as effective as thallium-201 for detecting exercise induced myocardial ischaemia. However, technetium-99m methoxy isobutylisonitrile provides a better image quality and may be a more sensitive marker of defect reversibility. For both radiotracers lung uptake is increased with extensive coronary artery disease and measurement of this variable provides prognostic information.