Current expert-recommended views for coronary angiography are based on heuristic experience and have not been scientifically studied. We sought to identify optimal viewing regions for first and second order vessel segments of the coronary arteries that provide optimal diagnostic value in terms of minimizing vessel foreshortening and overlap. Using orthogonal 2D images of the coronary tree, 3D models were created from which patient-specific optimal view maps (OVM) allowing quantitative assessment of vessel foreshortening and overlap were generated. Using a novel methodology that averages 3D-based optimal projection geometries, a universal OVM was created for each individual coronary vessel segment that minimized both vessel foreshortening and overlap. A universal OVM model for each coronary segment was generated based on data from 137 patients undergoing coronary angiography. We identified viewing regions for each vessel segment achieving a mean vessel foreshortening value of 5.8 +/- 3.9% for the left coronary artery (LCA) and 5.6 +/- 3.6% for the right coronary artery (RCA). The overall mean overlap values achieved were 8.7 +/- 7.9% for the LCA and 4.6 +/- 3.2% for the RCA. This scientifically-based OVM evaluation of coronary vessel segments provides the means to facilitate acquisitions during coronary angiography and interventions that minimize imaging inaccuracies related to foreshortening and overlap, improving the accuracy, efficiency, and safety of diagnostic and interventional coronary procedures.