Hemodynamics is widely believed to influence coronary artery bypass graft (CABG) stenosis. Although distal anastomosis has been extensively investigated, further studies on proximal anastomosis are still necessary, as the extent and initiation of the stenosis process may be influenced by the flow of the proximal anastomosis per se. Therefore, in this study, two models (i.e. 90° and 135° anastomotic models) were designed and constructed to simulate a proximal anastomosis of CABG for the left and right coronary arteries, respectively. Flow characteristics for these models were studied experimentally in order to validate the simulation results found earlier. PIV measurements were carried out on two Pyrex glass models, so that the disturbed flow (stagnation point, flow separation and vortex) found in both proximal anastomosis models using numerical simulation, could be verified. Consequently, a fair agreement between numerical and experimental data was observed in terms of flow characteristics, velocity profiles and wall shear stress (WSS) distributions under both steady and pulsatile flow conditions. The discrepancy was postulated to be due to the difference in detailed geometry of the physical and computational models, due to manufacturing limitations. It was not possible to reproduce the exact shape of the computational model when making the Pyrex glass model. The analysis of the hemodynamic parameters based on the numerical simulation study also suggested that the 135° proximal anastomosis model would alleviate the potential of intimal thickening and/or atherosclerosis, more than that of a 90° proximal anastomosis model, as it had a lower variation range of time-averaged WSS and the lower segmental average of WSSG.