P23H is the most common mutation in the RHODOPSIN (RHO) gene leading to a dominant form of retinitis pigmentosa (RP), a rod photoreceptor degeneration that invariably causes vision loss. Specific disruption of the disease P23H RHO mutant while preserving the wild-type (WT) functional allele would be an invaluable therapy for this disease. However, various technologies tested in the past failed to achieve effective changes and consequently therapeutic benefits. We validated a CRISPR/Cas9 strategy to specifically inactivate the P23H RHO mutant, while preserving the WT allele in vitro. We, then, translated this approach in vivo by delivering the CRISPR/Cas9 components in murine Rho +/P23H mutant retinae. Targeted retinae presented a high rate of cleavage in the P23H but not WT Rho allele. This gene manipulation was sufficient to slow photoreceptor degeneration and improve retinal functions. To improve the translational potential of our approach, we tested intravitreal delivery of this system by means of adeno-associated viruses (AAVs). To this purpose, the employment of the AAV9-PHP.B resulted the most effective in disrupting the P23H Rho mutant. Finally, this approach was translated successfully in human cells engineered with the homozygous P23H RHO gene mutation. Overall, this is a significant proof-of-concept that gene allele specific targeting by CRISPR/Cas9 technology is specific and efficient and represents an unprecedented tool for treating RP and more broadly dominant genetic human disorders affecting the eye, as well as other tissues. © The Author(s) 2017. Published by Oxford University Press. All rights reserved.