Myocardium lacks the ability to regenerate following injury. This is in contrast to skeletal muscle (SKM), in which capacity for tissue repair is attributed to the presence of satellite cells. It was hypothesized that SKM satellite cells multiplied in vitro could be used to repair injured heart muscle. Fourteen dogs underwent explantation of the anterior tibialis muscle. Satellite cells were multiplied in vitro and their nuclei were labelled with tritiated thymidine 24 h prior to implantation. The same dogs were then subjected successfully to a myocardial injury by the application of a cryoprobe. The cells were suspended in serum-free growth medium and autotransplanted within the damaged muscle. Medium without cells was injected into an adjacent site to serve as a control. Endpoints comprised histology using standard stains as well as Masson trichrome (specific for connective tissue), and radioautography. In five dogs, satellite cell isolation, culture, and implantation were technically satisfactory. In three implanted dogs, specimens were taken within 6-8 wk. There were persistence of the implantation channels in the experimental sites when compared to the controls. Macroscopically, muscle tissue completely surrounded by scar tissue could be seen. Masson trichrome staining showed homogeneous scar in the control site, but not in the test site where a patch of muscle fibres containing intercalated discs (characteristic of myocardial tissue) was observed. In two other dogs, specimens were taken at 14 wk postimplantation. Muscle tissue could not be found. These preliminary results could be consistent with the hypothesis that SKM satellite cells can form neo-myocardium within an appropriate environment. Our specimens failed to demonstrate the presence of myocyte nuclei. It is therefore further hypothesized that in the late postoperative period, the muscle regenerate failed to survive.