Asynchronicity in echocardiographic images is normally assessed visually. No prior quantitative studies have determined the limitations of this approach. To quantify visual recognition of myocardial asynchronicity in echocardiographic images, computer-simulated delay phantom loops were generated from a 3.3 MHz digital image data from a normal left ventricular short-axis heart cycle acquired at 55 frames per second. Six expert observers visually assessed 30 abnormal and 3 normal loops with differing computer-induced delay patterns on 3 occasions and in this optimally simulated environment could recognize only single delays of 89 ms or more. This was improved to 71 ms or more by use of side-by-side (normal versus abnormal) comparative review. Thus visual assessment of clinically important regional delay in rest or stress echo images is limited.