Risk stratification of coronary artery disease may provide a basis for selection of treatment to prevent myocardial events and to assist functional recovery. Iodine-123 (rho-iodophenyl)-3-R,S-methylpentadecanoic acid (123I-BMIPP) is a radioiodinated fatty acid analogue for single-photon emission tomographic (SPET) imaging, and several reports have demonstrated that the abnormal uptake of 123I-BMIPP is associated with wall motion abnormality and severe coronary artery stenosis. Clarification of the contribution of fatty acids to myocardial metabolism would be highly valuable in recognising this critical condition. In this study, we investigated the myocardial uptake of 123I-BMIPP under low-flow ischaemia, and compared it with the uptake of fluorine-18 fluorodeoxyglucose (18F-FDG). Using open chest dogs, the flow of the left anterior descending coronary artery was controlled using a pneumatic occluder in order to maintain a 30%-40% reduction of Doppler flow. 123I-BMIPP and 18F-FDG were injected into the left atrium after 90 min of ischaemia (protocols 1 and 3). Canine hearts were excised after 120 min of ischaemia for the measurement of radioactivity. In protocol 2, 123I-BMIPP alone was injected and hearts were excised 8 min after the injection. A time-course biopsy study was also performed at the same time (protocol 3). Wall thickening was evaluated using a wall tracker module. The uptake of 18F-FDG increased significantly in the ischaemic region (232%+/-135% vs non-ischaemic, P<0.05 in protocol 1) even on mild reduction of myocardial blood flow (MBF). The increased uptake of 18F-FDG did not correlate well with the severity of MBF. On the other hand, 123I-BMIPP uptake decreased gradually (78.9%+/-23.6%, P<0.05 in protocol 1, and 85.9%+/-24.3% in protocol 2) in the ischaemic region, specifically in the endocardium (64.0%+/-28.9%, P<0.05 in protocol 1, and 75.1%+/-28.8%, P<0.05 in protocol 2), and correlated strongly with MBF (r=0.93 in protocol 1 and r=0.97 in protocol 2) as a logarithmic function. This indicated that the abnormal uptake of 123I-BMIPP was associated not only with wall motion abnormality but also with the severity of MBF. In the biopsy study (protocol 3), the radioactivity of either 123I-BMIPP or 18F-FDG correlated well with the MBF at the time of tracer injection and was similar to post-mortem analysis. It is concluded that 18F-FDG is a valid tool for identifying ischaemic myocardium even in its earliest stages. On the other hand, 123I-BMIPP might be used to detect moderately to severely ischaemic myocardium such as hibernation, suggesting the potential value of 123I-BMIPP in the risk stratification of patients with severe coronary artery disease who require revascularisation without delay.