Optic nerve (ON) injury causes blindness in adult mammals as their retinal ganglion cells (RGCs) cannot regenerate axons. However, amphibian RGC axons do not experience the same regenerative failure. Studying the regeneration process of the ON in amphibians holds profound implications for regenerative medicine and human health. Using transgenic tadpoles and laser micro-optics, we developed a reproducible ON transection and regeneration model. Through microscopy of axon dynamics, functional testing to assess visual pathway recovery, TUNEL cell death and EdU cell proliferation assays, and RNA-seq of the retina and optic nerve, we characterized the optic nerve injury response and subsequent recovery. Our model suggests no chemoattractant gradient exists early in regeneration, with defasciculated axons sprouting in random directions from the globe-proximal cut end. Once individual axons reach the appropriate targets in the brain, their tract is reinforced by other regenerating axons, restoring normal ON morphology. Thus, guidance cues or scaffolding from brain-innervating axons likely support later stages of regeneration. After 14 days, the regenerated ON is morphologically indistinguishable from the naïve ON, and visual function is restored. We found no evidence of RGC death or new RGC formation in the model, suggesting that ON regeneration involves remodeling of injured axons of pre-existing RGCs.