This study looks at blood flow in four different human right coronary arteries (RCAs), which have been reconstructed from bi-plane angiograms. A generalized power-law model of blood viscosity is used to study the blood flow at a particular point in the cardiac cycle. Large differences are found in the wall shear stress magnitude (WSS) distributions in the four arteries, leading to the conclusion that it is not possible to make generalizations based on the study of a single artery. The pattern of WSS is found to be related to the geometry of a particular artery, that is, lumen diameter and arterial curvature as well as a combination of these two factors. There is a strong correlation between WSS and reciprocal radius and a weaker correlation between high curvature and extremes of WSS, with high WSS on the 'inside' of a bend and low WSS on the 'outside' of a bend. This is in contrast to the situation for a simple curved tube with constant radius where the inverse is observed. For each artery, a region proximal to the acute margin is identified where low WSS is found and where WSS is lower on the 'inner' wall of the RCA than on the 'outer' wall. This region is one where several studies have found that the human RCA preferentially exhibits atherogenesis.