High-frequency epicardial echocardiography (HFE) is an ultrasound imaging technique capable of visualizing coronary arteries in cross-section. Use of this imaging technique in the intra-operative setting can provide information about the structure of coronary arteries. As a result of time constraints imposed in this setting, there is a need for rapid, on-line analysis of HFE images. An algorithm for segmenting HFE images of coronary arteries into wall and luminal regions is described in this paper. The algorithm is more objective and time efficient than manual methods. A validation study with post-mortem human hearts demonstrated an excellent correlation (r = 0.99 for luminal areas, r = 0.99 for wall areas) between HFE images segmented with the algorithm and by manual methods, and a good correlation (r = 0.90 for luminal areas, r = 0.86 for wall areas) between HFE images segmented with the algorithm and manually segmented images of histological samples of the corresponding coronary arterial segments. Use of the algorithm for intra-operative HFE image segmentation may reduce the amount of time required for HFE image analysis and allow for an increased amount of data collection and analysis in the operating room.