BACKGROUND: There is nearly no muscle tissue with satisfactory function and appearance applying in clinical repair and construction of injured muscles to date. OBJECTIVE: To investigate the feasibility of applying autologous fascia as a scaffold to construct muscle in vivo. DESIGN, TIME AND SETTING: The randomized, self-matched control experiment was carried out between January 2004and June 2006 at Department of Burns & Plastic Surgery, Second Hospital of Shenzhen, Shenzhen, Guangdong Province, China MATERIALS: Twenty-eight healthy New Zealand rabbits, weighing (1.7±0.5) kg, without sex restriction, establishing middle part defect model of anterior tibial muscle of rabbit hind legs. METHODS: One hind leg of each rabbit was randomized to the experimental group (n=28), the other hind leg was assigned to one of 3 control groups, scaffold-connected group (n=10), muscle particle implant group (n=10) and blank control group(n=8). In experimental group, the defect was connected with an autologous fascial scaffold and filled with the mutilated muscle particles, and subcutaneous tissue and skin were sutured in situ. In scaffold-connected group, the treatments were same to the experimental group only except muscle particle implantation. In muscle particle implant group, the defect was filled with muscle particles but without fascial scaffold and other treatments were same to the experimental group. The defect in blank control group received no treatment. MAIN OUTCOME MEASURES: The success rate of muscle transplantation, histological and ultra structural observation,and immunohistochemical identification of desmin were observed at 2, 3, 4, 6 and 9 weeks after operation. The middle parts of samples were also harvested for relative quantitative analysis of α-actin cDNA using reverse transcription-polymerase chain reaction in the experimental group and scaffold-connected group.RESULTS: In experimental group, 1 muscle broke near the proximal junction, the other 27 succeeded and the appearance of healed defects became near normal gradually. In scaffold-connected group, 4 muscles broke, 6 muscles still depressed in defect area; in muscle particle implant group and blank control group, the defects had no change. In experimental group, a large quantity of skeletal muscle satellite cells proliferated, which reached peak at 2-3 weeks, cells attached to the ends of fibrous connective tissue; in scaffold-connected group only fibrous connective tissue was seen. lmmunohistochemistry showed that 85% cells in experimental group were desmin-positive, while the positive rate in scaffold-connected group was < 25%. The relative quantitative analysis of α -actin cDNA showed that there were significant differences between the experimental group and scaffold-connected group at different time points(P < 0.05).CONCLUSION: The success rate of repairing muscle defect with autologous fascial scaffold reached 93.33%, which indicates that it is feasible to promote muscle regeneration with autologous fascial scaffold.