Background: Blood oxygen level dependent (BOLD) T2* MRI detects signal variance within the myocardium based on changes in the deoxyhaemoglobin level following pharmacological stress, and it has the potential to identify areas of myocardial ischemia. The aim of the present study was to assess the utility of BOLD T2* MRI in the detection of myocardial ischemia in patients with an existing diagnosis of coronary artery disease. Method: Twenty-one patients with established three-vessel coronary artery disease on coronary angiography underwent rest and dipyridamole stress MRI using a double breath-hold T2* weighted ECG gated sequence. Analysis was performed on multiple short-axis slices of the heart, projected as a bull’s eye. The myocardium was divided into three coronary territories, yielding 63 territories in total. A signal change between rest and stress of more than +/- 4% was significant, implying a change in deoxyhaemoglobin concentration. A signal decrease or no changes denote the presence of ischemia, while a signal increase indicates no ischemia. Results: All images were of sufficient quality for signal intensity analysis. In 12/63 territories (19%), a significant signal increase following stress was detected. A significant signal decrease was detected in 34/63 territories (54%), and in 17/63 territories (27%) there was a nonsignificant change. The presence of a perfusion defect was identified, therefore, in 51/63 (81%), based on the signal difference between rest and stress. Conclusion: Changes in myocardial oxygen level appear to be detectable by BOLD T2* MRI without using contrast media. Further, larger comparative studies are required to evaluate the diagnostic and prognostic impact of this technique and to compare it to the gold standard methods for the detection of myocardial ischemia. (C) 2006 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.