Running is believed to be beneficial for human health. Many studies have focused on the neuroprotective effects of voluntary running on animal models. There were both primary and secondary degeneration in neurodegenerative diseases, including glaucoma. However, whether running can delay primary or sec-ondary degeneration or both of them was not clear. Partial optic nerve transection model is a valuable glaucoma model for studying both primary and secondary degeneration because it can separate primary (mainly in the superior retina) from secondary (mainly in the inferior retina) degeneration. Therefore, we compared the survival of retinal ganglion cells between Sprague-Dawley rat runners and non-runners both in the superior and inferior retinas. Excitotoxicity, oxidative stress, and apoptosis are involved in the degeneration of retinal ganglion cells in glaucoma. So we also used west immunoblotting to compare the expression of some proteins involved in apoptosis (phospho-c-Jun N-terminal kinases, p-JNKs), oxida-tive stress (manganese superoxide dismutase, MnSOD) and excitotoxicity (glutamine synthetase) between runners and non-runners after partial optic nerve transection. Results showed that voluntary running delayed the death of retinal ganglion cells vulnerable to primary degeneration but not those to secondary degeneration. In addition, voluntary running decreased the expression of glutamine synthetase, but not the expression of p-JNKs and MnSOD in the superior retina after partial optic nerve transection. These results illustrated that primary degeneration of retinal ganglion cells might be mainly related with excitotoxicity rather than oxidative stress; and the voluntary running could down-regulate excitotoxicity to delay the pri-mary degeneration of retinal ganglion cells after partial optic nerve transection.