Oxygen tension (PO₂) of urine in the bladder could be used to monitor risk of acute kidney injury if it varies with medullary PO₂. Therefore, we examined this relationship, and characterized oxygen diffusion across walls of the ureter and bladder, in anesthetized rabbits. A computational model was then developed to predict medullary PO2 from bladder urine PO₂. Both intravenous infusion of [Phe²,Ile³,Orn⁸]-vasopressin and infusion of N^G-nitro-L-arginine reduced urinary PO₂ and medullary PO₂ (8-17%), yet had opposite effects on renal blood flow and urine flow. Changes in bladder urine PO₂ during these stimuli correlated strongly with changes in medullary PO₂ (within-rabbit r² = 0.87-0.90). Differences in the PO₂ of saline infused into the ureter close to the kidney could be detected in the bladder, although this was diminished at lesser ureteric flow. Diffusion of oxygen across the wall of the bladder was very slow, so was not considered in the computational model. The model predicts PO₂ in the pelvic ureter (presumed to reflect medullary PO₂) from known values of bladder urine PO₂, urine flow and arterial PO₂. Simulations suggest that, across a physiological range of (single kidney) urine flow in anesthetized rabbits (0.1 - 0.5 ml/min), a change in bladder urine PO₂ reflects 10 - 50% of the change in pelvic urine/medullary PO₂. Thus, it is possible to infer changes in medullary PO₂ from changes in urinary PO₂, so urinary PO₂ may have utility as a biomarker of risk of acute kidney injury.