Recessive dystrophic epidermolysis bullosa (RDEB) is a debilitating and ultimately lethal blistering disease caused by mutations to the Col7a1 gene. Development of novel cell therapies for the treatment of RDEB would be fostered by having immunodeficient mouse models able to accept human cell grafts; however, immunodeficient models of many genodermatoses such as RDEB are lacking. To overcome this limitation, we combined the clustered regularly interspaced short palindromic repeats and associated nuclease (CRISPR/Cas9) system with microinjection into NOD/SCID IL2rγc null (NSG) embryos to rapidly develop an immunodeficient Col7a1−/− mouse model of RDEB. Through dose optimization, we achieve F0 biallelic knockout efficiencies exceeding 80%, allowing us to quickly generate large numbers of RDEB NSG mice for experimental use. Using this strategy, we clearly demonstrate important strain-specific differences in RDEB pathology that could underlie discordant results observed between independent studies and establish the utility of this system in proof-of-concept human cellular transplantation experiments. Importantly, we uncover the ability of a recently identified skin resident immunomodulatory dermal mesenchymal stem cell marked by ABCB5 to reduce RDEB pathology and markedly extend the lifespan of RDEB NSG mice via reduced skin infiltration of inflammatory myeloid derivatives.