External detection by a single collimator of the washout curve of 133Xe following its intracoronary injection was used in humans following its validation in animals. However, additional experimental studies showed that the straightforward application of the theoretical principles on which the technique should be based resulted in uncorrected values. Special empirical modifications in the analysis of the washout curves were required in order to obtain a fair agreement between calculated and experimental flow values. Furthermore, studies in humans indicate that reappearance of indicator into the counting field due to recirculation and diffusion retention of the tracer in epicardial fat alter the precordial washout curve leading to underestimation of coronary blood flow. This method has found a second wave of applicatin in humans for the study of regional myocardial perfusion using a gamma camera. The amount of new information on regional myocardial perfusion provided by this approach is quite considerable; however, the interpretation of the values obtained in these studies requires particular attention in view of the limitations outlined above.