The aim of this study was to evaluate how the diagnostic accuracy of a stress echocardiographic procedure, such as a dipyridamole echocardiography test, depends on the specific experience of the physician interpreting the test. Recordings of 50 consecutive dipyridamole echocardiographic tests were selected for the first part of the study. They were analyzed by 20 experienced echocardiographers with different backgrounds in stress echocardiography: 10 beginners (less than 20 stress studies interpreted with trained staff) and 10 experienced observers (greater than or equal to 100 stress studies performed). Diagnostic accuracy (true positive + true negative/total number of tests) versus the angiographic reference standard (greater than 70% coronary stenosis of at least one major coronary artery) was 62 +/- 6% for beginners and 85 +/- 3% for experienced observers (p less than 0.0001). In the second part of the study, 10 observers (5 beginners and 5 experienced observers) evaluated 2 different sets of 50 dipyridamole echocardiographic test studies before and after the training of the beginners. Before training, the accuracy of beginners was lower than that of experienced observers (61 +/- 7% versus 85 +/- 3%; p less than 0.001). After training, the accuracy gap was closed (83 +/- 3% versus 86 +/- 2%; p = NS). Therefore, interpretation of stress echocardiographic tests by an echocardiographer without specific training severely underestimates the diagnostic potential of this technique. One hundred stress echocardiographic studies are more than adequate to build the individual learning curve and reach the plateau of diagnostic accuracy that the test can yield.