Dobutamine stress echocardiography (DSE) has been shown to be a very useful non-invasive technique for the detection of coronary artery disease. However, inadequate transthoracic images preclude the use of DSE in a significant proportion of patients. Transesophageal (TEE) or transthoracic contrast echocardiography (CE) can however overcome this limitation. The comparison between the two techniques has never been investigated during a stress test. Therefore, we designed a prospective study to compare DSE-CE and DSE-TEE for the detection of coronary artery disease in patients with poor echo image quality. We studied 42 patients scheduled for quantitative coronary angiography. Prospective DSE-CE and DSE-TEE with maximum one day interval were performed in a random order. Significant coronary artery disease was detected in 30 patients, nine with single vessel disease and 21 with multivessel disease. Sensitivity of DSE was higher with CE than with TEE (90% vs 87%, p=NS). There was no significant difference with respect to specificity in both groups (100% vs 92%, p=NS). The diagnostic accuracy was similar in both groups (93% vs 88%, NS). The kappa value for identical interpretation of a stress echocardiography study was nearly identical with both modalities 0.75 to 0.78. In poorly echogenic patients, DSE-CE is a valuable alternative for the detection of myocardial ischemia in comparison with DSE-TEE. Because DSE-CE is more comfortable than TEE, it should be used in patients with suboptimal transthoracic echocardiograms for the evaluation of coronary artery disease during DSE.