This study tested the hypothesis that endurance training initiated 28 d following grafting of the soleus muscle would increase fiber cross-sectional area concomitant with an increase in mass. Nerve-implant orthotopic grafting operations were performed on 6-wk-old rats anesthetized with pentobarbital sodium. A cohort of animals began running 28 d later. Control muscles were from age-matched, untrained rats. Mass and fiber cross-sectional area of grafts were 37 and 66% less than the respective control values at both 56 and 112 d post-grafting. Training increased graft mass by 49% over the non-run graft value of 82 +/- 8 mg at 56 d post-grafting. Continued training did not increase mass further. The grafts of trained rats were 33% greater than untrained at day 112 due to growth of grafts in untrained rats. Running had no effect on fiber cross-sectional area of grafts through 56 d, but by 112 d the cross-sectional area of Type I fibers was 30% greater than the non-run graft value of 1,271 +/- 81 micron2. By 112 d fiber type profiles were not different between control muscle and grafts from trained and untrained rats. We conclude that there is a dissociation between mass and fiber cross-sectional area in grafts compared to control muscle, and training affects these variables by similar magnitudes but at different times.