Treatment effects observed in animal studies often fail to be recapitulated in clinical trials. While this problem is multifaceted, one reason for this failure is the use of inadequate laboratory models. It is challenging to model complex human diseases in traditional laboratory organisms, but this issue can be circumvented through the study of human xenografts. The surgical method we describe here allows for the creation of human skeletal muscle xenografts, which can be used to model muscle disease and to carry out preclinical therapeutic testing. Under an Institutional Review Board (IRB)-approved protocol, skeletal muscle specimens are acquired from patients and then transplanted into NOD-Rag1null IL2rγnull (NRG) host mice. These mice are ideal hosts for transplantation studies due to their inability to make mature lymphocytes and are thus unable to develop cell-mediated and humoral adaptive immune responses. Host mice are anaesthetized with isoflurane, and the mouse tibialis anterior and extensor digitorum longus muscles are removed. A piece of human muscle is then placed in the empty tibial compartment and sutured to the proximal and distal tendons of the peroneus longus muscle. The xenografted muscle is spontaneously vascularized and innervated by the mouse host, resulting in robustly regenerated human muscle that can serve as a model for preclinical studies.