BACKGROUND--Coronary artery imaging is an important investigation for the management of coronary artery disease. The only reliable technique presently available, x ray contrast angiography, is invasive and is associated with a small morbidity and mortality. Alternative non-invasive imaging would be useful, but the small calibre and tortuosity of the coronary vessels, and cardiac and respiratory motion create formidable imaging problems. OBJECTIVE--The development of rapid magnetic resonance imaging of the coronary arteries. PATIENTS--21 healthy controls and five patients with coronary artery disease established by x ray contrast angiography, of whom two had undergone bypass grafting. METHODS--Magnetic resonance imaging was performed with gradient echoes and a segmented k-space technique, such that a complete image was acquired in 16 cardiac cycles during a breathhold. The signal from fat was suppressed and images were acquired in late diastole to reduce artefact from cardiac motion. An imaging strategy was developed for the proximal arteries, including longitudinal imaging from oblique planes defined according to the origins and the continuation of the arteries in the atrioventricular grooves or interventricular sulcus. RESULTS--Of the 26 subjects studied, 22 were imaged successfully. Identification of the artery was possible for the left main stem, left anterior descending, right coronary, and left circumflex arteries respectively in 95%, 91%, 95%, and 76%. The arterial diameter at the origin could be measured in 77%, 77%, 81%, and 63%. The mean (SD) arterial diameter in each case (4.8 (0.8), 3.7 (0.5), 3.9 (0.9), and 2.9 (0.6) mm) was not significantly different from reference values. The mean length of artery visualised was 10.4 (5.2), 46.7 (22.8), 53.7 (27.9), and 26.3 (17.5) mm. In 12 healthy men the total coronary area was 30.9 (9.2) mm2 and the ratio compared with body surface area was 16.4 (4.4) mm2m2 (both p = NS compared with reference values). In seven patients in whom x ray contrast coronary angiography was available, the proximal arterial diameter was 3.9 (1.1) mm measured by magnetic resonance and 3.7 (1.0) mm by x ray contrast angiography (p = NS). The mean difference between the measurements was 0.2 (0.5) mm, and the coefficient of variation was 13.7%. All five occluded coronary arteries were identified, as were all three vein grafts. In two patients insertion of the graft into the native arteries was identified. CONCLUSIONS--Magnetic resonance coronary angiography is feasible. Good results were obtained by a breath-hold, fat suppression technique, gated to late diastole. Arterial occlusions and vein grafts were readily identified. Further studies are required to establish its value in the detection of coronary stenosis and to develop the measurement of coronary flow velocity which could be used to quantify the severity of the stenosis.