Rhabdomyosarcoma (RMS) is the most frequent form of paediatric soft-tissue sarcoma and remains a medical challenge, holding in failure current therapeutic strategies. RMS shares histological features with cells of the muscle lineage and this cancer is thought to arise from malignant transformation of myogenic precursors. It has been proposed that RMS and myogenesis could represent the Jekyll and Hyde of skeletal muscle. The underlying idea is that some early steps of myogenic differentiation are blocked in RMS, and that understanding how the normal process has gone awry could help to decipher the biological underpinnings of tumorigenesis and tumor escape. It is widely agreed that extracellular matrix (ECM) interferes in skeletal muscle regeneration and that defects in ECM components are clinically significant. The challenge is now to decipher actors and mechanisms responsible for the transmission of signals to muscle cells and the subsequent alterations that could be associated with RMS. Using an original transgenic mice model, we show here that ADAMTSL1 is involved in skeletal muscle regeneration. As previously reported for other members of its family, ADAMTSL1 is part of the TGF-b-ECM-sequestering complex and likely positively regulates TGF-b-pathway activity. Last, we observed that ADAMTSL1 expression behaves as a strong prognostic factor in the aggressive fusion-positive RMS and correlates with a neural-like phenotype of tumor cells, resulting from gain of SMAD2/3/4 targets.