Electrical stimulation of the retina can elicit visual perceptions in patients affected by certain types of retinal degeneration by activating some of the surviving neural cells. Even though retinal prostheses are now commercially available, many conceptual and technological issues remain to be clarified. Computational modelling has contributed in the past to optimize peripheral nerve, spinal cord, and deep brain electrical stimulation, and has the potential to provide a better understanding of how the retina responds to electrical stimulation, to improve the operation of existing implants, or to propose alternative designs. In this thesis, we developed a new computational framework to study retinal electrical stimulation.