The renal medulla is prone to hypoxia. Medullary hypoxia is postulated to be a leading cause of acute kidney injury, so there is considerable interest in predicting the oxygen tension in the medulla. Therefore we have developed a computational model for blood and oxygen transport within a physiologically normal rat renal medulla, using a multilevel modeling approach. For the top-level model we use the theory of porous media and advection-dispersion transport through a realistic three-dimensional representation of the medulla's gross anatomy to describe blood flow and oxygen transport throughout the renal medulla. For the lower-level models, we employ two-dimensional reaction-diffusion models describing the distribution of oxygen through tissue surrounding the vasculature. Steady-state model predictions at the two levels are satisfied simultaneously, through iteration between the levels. The computational model was validated by simulating eight sets of experimental data regarding renal oxygenation in rats (using 4 sets of control groups and 4 sets of treatment groups, described in 4 independent publications). Predicted medullary tissue oxygen tension or microvascular oxygen tension for control groups and for treatment groups that underwent moderate perturbation in hemodynamic and renal functions is within ±2 SE values observed experimentally. Diffusive shunting between descending and ascending vasa recta is predicted to be only 3% of the oxygen delivered. The validation tests confirm that the computational model is robust and capable of capturing the behavior of renal medullary oxygenation in both normal and early-stage pathological states in the rat.