The retina's high oxygen demands and the retinal vasculature's relatively sparse nature are assumed to contribute to the retina's specific vulnerability to vascular diseases. This study has been designed to model the oxygen transport in physiologically realistic retinal networks. A computational fluid dynamics study has been conducted to investigate the effect of topological changes on the oxygen partial pressure distribution in retinal blood vessels. The Navier Stokes equations for blood flow and the mass transport equation for oxygen have been coupled and solved simultaneously for the laminar flow mass transfer problem. The mean oxygen saturation of a healthy eye has been 93% in retinal arterioles and 58% in venules. The arteriovenous difference has been 35%. For a patient with a central retinal vein occlusion (CRVO), the mean oxygen saturation has been 33%. The findings from the analysis are generally consistent with a lot of previous experimental measurements and clinical data available in the literature, demonstrating the efficiency of our model for predicting the oxygen distribution in the retinal networks. This paves the way for a new research and applications for simulating inaccessible cases from experimental studies.