According to the Stewart-Hamilton equation flow is inversely related to the area under the time-concentration curve produced by the transit of a detectable indicator. To verify the applicability of this principle for contrast echocardiography, we bolus injected a saccharide echo contrast agent (0.8 ml) into an in vitro circulatory model at variable flow rates. Two-dimensional echo images were digitized, and curves demonstrating the ratio of videointensity over time were derived. As expected, flow was inversely related to the area under the curves (r = 0.93). To apply this principle to human coronary circulation, we bolus injected sonicated iopamidol (4 ml) into the normal left coronary artery of six patients at baseline and after intravenous administration of dipyridamole (0.84 mg/kg in 10 minutes). Echo images were digitized, and myocardial time-intensity curves were derived. The area under the curve after dipyridamole administration (210 +/- 128 gray level.sec) did not appear significantly different from that at baseline (177 +/- 80 gray level.sec). Thus a mismatch exists between contrast echo data obtained in vitro and in human coronary circulation.