It is unclear whether echocardiography at peak bicycle exercise adds information to registrations obtained recumbent immediately after the test and what factors influence image quality. Therefore, exercise echocardiography was performed consecutively and prospectively in 66 men, unselected with regard to echocardiography, one month after an episode of unstable coronary artery disease. Of 594 segments (9 x 66), 569 (96%) were adequately visualized recumbent at rest. The corresponding figures recumbent directly after exercise, seated before exercise, and seated at peak exercise were 544 (92%), 474 (80%), and 428 (72%), respectively. In the majority of our patients, acceptable images at peak exercise were obtained for the septal region, while for the anterior, lateral, and inferior segments the success rate varied from 50 to 70%. Recumbent after exercise, the success rate was acceptable for most segments, possibly with the exception of the apical and lateral segments. Fifty-five patients developed new wall motion abnormalities or worsening of wall motion in connection with exercise. Echocardiography at peak exercise provided more information than afterwards in patients with images of good quality. However, in patients with inferior image quality, the registrations obtained recumbent after the test revealed wall motion abnormalities which were not obtained seated at peak exercise. Patients with worse image quality had significantly higher respiratory rate and weight, and rated a higher degree of dyspnoea at peak exercise than those with good quality. We conclude that in middle aged men with coronary artery disease, image acquisition at peak bicycle exercise and immediately after exercise are of complementary value.