It was hypothesized that regional myocardial blood flow could be measured using myocardial contrast echocardiography. Accordingly, arterial blood was perfused into the coronary circulation in 16 dogs. In Group 1 dogs (n = 8), blood flow to the cannulated left circumflex artery was controlled with use of a roller pump, whereas in Group 2 dogs (n = 8) blood flow to the left anterior descending coronary artery was controlled by a hydraulic occluder placed around it. Sonicated microbubbles (mean size 4 microns) were used as the contrast agent. In Group 1 dogs the microbubbles were injected subselectively into the left circumflex artery, whereas in Group 2 dogs they were injected selectively into the left main coronary artery and two-dimensional echocardiographic images were recorded. Computer-generated time-intensity curves were derived from these images and variables of these curves correlated with transmural blood flow measured with radiolabeled microspheres. A gamma-variate function (y = Ate-alpha t) best described the curves, and alpha (a variable of curve width) correlated well with transmural blood flow at different flow rates in all Group 1 and Group 2 dogs (mean r = 0.81 and 0.97, respectively). Other variables of the curve width also correlated well with myocardial blood flow, but peak intensity had a poor correlation with myocardial blood flow in both groups of dogs (r = 0.39 and r = 0.63, respectively). When data from all dogs were pooled, Group 1 dogs still showed good correlation between variables of curve width and myocardial blood flow (r = 0.81); Group 2 dogs did not (r = 0.45). The difference between the two sets of dogs was related to the site of contrast agent injection. It is concluded that measurement of the transit time of microbubbles through the myocardium with two-dimensional echocardiography accurately reflects regional myocardial blood flow. Although injection of contrast agent selectively into the left main coronary artery only allows measurement of relative flow, it may be feasible to measure absolute flow by injecting contrast agent subselectively into a coronary artery. Myocardial contrast echocardiography may, therefore, offer the unique opportunity of simultaneously assessing regional myocardial perfusion and function in vivo.