The oxygen tension of the optic nerve is regulated by the intraocular pressure and systemic blood pressure, the resistance in the blood vessels and oxygen consumption of the tissue. The oxygen tension is autoregulated and moderate changes in intraocular pressure or blood pressure do not affect the optic nerve oxygen tension. If the intraocular pressure is increased above 40 mmHg or the ocular perfusion pressure decreased below 50 mmHg the autoregulation is overwhelmed and the optic nerve becomes hypoxic. A disturbance in oxidative metabolism in the cytochromes of the optic nerve can be seen at similar levels of perfusion pressure. The levels of perfusion pressure that lead to optic nerve hypoxia in the laboratory correspond remarkably well to the levels that increase the risk of glaucomatous optic nerve atrophy in human glaucoma patients. The risk for progressive optic nerve atrophy in human glaucoma patients is six times higher at a perfusion pressure of 30 mmHg, which corresponds to a level where the optic nerve is hypoxic in experimental animals, as compared to perfusion pressure levels above 50 mmHg where the optic nerve is normoxic. Medical intervention can affect optic nerve oxygen tension. Lowering the intraocular pressure tends to increase the optic nerve oxygen tension, even though this effect may be masked by the autoregulation when the optic nerve oxygen tension and perfusion pressure is in the normal range. Carbonic anhydrase inhibitors increase the optic nerve oxygen tension through a mechanism of vasodilatation and lowering of the intraocular pressure. Carbonic anhydrase inhibition reduces the removal of CO2 from the tissue and the CO2 accumulation induces vasodilatation resulting in increased blood flow and improved oxygen supply. This effect is inhibited by the cyclo-oxygenase inhibitor, indomethacin, which indicates that prostaglandin metabolism plays a role. Laboratory studies suggest that carbonic anhydrase inhibitors might be useful for medical treatment of optic nerve and retinal ischemia, potentially in diseases such as glaucoma and diabetic retinopathy. However, clinical trials and needed to test this hypotheses.