Abstract: The complex and mostly multiple and unknown aetiology of neurodegenerative diseases always give way to an intricate scenario of dying tissue that involves multiple cell mediators and cell types. All neurodegenerative diseases of the central nervous system (CNS) share common mechanisms, regardless their origin: oxidative stress, neuroinflammation and cell death. Accordingly, retinal degenerative diseases, with or without a genetic cause, as retinitis pigmentosa (RP), glaucoma, age-related macular degeneration (AMD) or diabetic retinopathy (DR) do not differ in their basic mechanisms of cell death neither one to another, nor from those observed in other CNS diseases as Parkinson’s or Alzheimer’s (Cuenca et al., 2014). Indeed, the therapeutic findings should be able to be more or less easily extrapolated between these conditions, as far as they are directed to common dartboards. Gene therapy, in which we have very high hopes to solve genetic disorders, is currently being traslated from preclinical assays to the clinic for some retinal degenerative diseases, with successful achievement up today for the Leber congenital amaurosis, due to mutations in the RPE65 gene (Garafalo et al., 2020). But our promising therapies still face relevant challenges. In this sense, CRISP/Cas editing tools used to amend genetic missenses, need to fix secondary effects, as those related to the immune response (Yu et al., 2017); stem cell approaches have to procure the functionality of transplanted cells in the recipient, to assure the accurate establishment of synaptic connectivity and cell contacts, and gain success in precise image processing (Cuevas et al., 2019; Garita-Hernandez et al., 2019); and optogenetics also needs to find appropriate vectors for the delivery and expression in suitable cell types, avoiding immunological rejection of the vector systems (Shen et al., 2020). While gene- and cell-based therapies evolve through the tortuous pathway of biological success, combined therapies with antioxidant (as lutein or zeaxanthin), antiinflammatory (as corticosteroids or cannabinoids), and antiapoptotic (as tauroursodeoxycholic acid or proinsulin) molecules appear currently as the widest approach to pharmacologically treat a wide spectrum of retinal degenerative diseases. These compounds provide several advantages. They can slow down the progression of the degenerative process, so preserving the visual capacity for a certain time. Moreover, the administration of neuroprotective factors is essential even when the vision has been completely lost, as they can improve non-visual functions, like the control of circadian rhythms and pupil contraction, as the cannabinoid-mediated improvement of circadian rhythmicity in P23H rats, which are mediated by the melanopsin-containing photosensitive ganglion cells (Lax et al., 2019). Non-visual retinal functions have also effects on memory and depression. Therefore, the preservation of this subset of cells, although will not improve the visual function, will surely improve the quality of life of the patients and should not be underestimated. But, far beyond, these molecules will surely increase the success of the new therapies, as they can provide an adequate environment of healthy cells, as a substrate for gene transplant or optogenetic approaches, which could hardly be successful in a damaged tissue surrounded by dying cells. Genetic material can be potentially incorporated to the retina and eventually restore the visual functionality in the zone in which it is injected but, without a global actuation on the whole retina by maintaining the health of the adjacent cells, an inflamed surrounding could end in a complete failure of any therapy. Hence, the concomitant use of antiinflammatory, antioxidant and antiapoptotic agents, as well as neurotrophic and growth factors, will provide an adequate environment of healthy cells that will help to achieve a sustained functional restoration of the visual function (Figure 1), as it has been shown for the combination of progesterone and lipoic acid in a mouse model of RP (Ramirez-Lamelas et al., 2018).