Stem cell transplantation is a promising potential therapy for muscular dystrophies, but for this purpose, the cells need to be systemically-deliverable, give rise to many muscle fibres and functionally reconstitute the satellite cell niche in the majority of the patient's skeletal muscles. Human skeletal muscle-derived pericytes have been shown to form muscle fibres after intra-arterial transplantation in dystrophin-deficient host mice. Our aim was to replicate and extend these promising findings. Isolation and maintenance of human muscle derived cells (mdcs) was performed as published for human pericytes. Mdscs were characterized by immunostaining, flow cytometry and RT-PCR; also, their ability to differentiate into myotubes in vitro and into muscle fibres in vivo was assayed. Despite minor differences between human mdcs and pericytes, mdscs contributed to muscle regeneration after intra-muscular injection in mdx nu/nu mice, the CD56+ sub-population being especially myogenic. However, in contrast to human pericytes delivered intra-arterially in mdx SCID hosts, mdscs did not contribute to muscle regeneration after systemic delivery in mdx nu/nu hosts. Our data complement and extend previous findings on human skeletal muscle-derived stem cells, and clearly indicate that further work is necessary to prepare pure cell populations from skeletal muscle that maintain their phenotype in culture and make a robust contribution to skeletal muscle regeneration after systemic delivery in dystrophic mouse models. Small differences in protocols, animal models or outcome measurements may be the reason for differences between our findings and previous data, but nonetheless underline the need for more detailed studies on muscle-derived stem cells and independent replication of results before use of such cells in clinical trials.