We tested the effects of prolonged voluntary wheel running on the muscle function of mdx mice treated with one of two different microdystrophin constructs. At 7 weeks of age mdx mice were injected with a single dose of AAV9-CK8-microdystrophin with (gene therapy 1, GT1) or without (gene therapy 2, GT2) the nNOS-binding domain and were assigned to one of four gene therapy treated groups: mdxRGT1 (run, GT1), mdxGT1 (no run, GT1), or mdxRGT2 (run,GT2), mdxGT2 (no run, GT2). There were two mdx untreated groups injected with excipient: mdxR (run, no gene therapy) and mdx (no run, no gene therapy). A third no treatment group, Wildtype (WT) received no injection and did not run. mdxRGT1, mdxRGT2 and mdxR performed voluntary wheel running for 52 weeks; WT and remaining mdx groups were cage active. Robust expression of microdystrophin occurred in diaphragm, quadriceps, and heart muscles of all treated mice. Dystrophic muscle pathology was high in diaphragms of non-treated mdx and mdxR mice and improved in all treated groups. Endurance capacity was rescued by both voluntary wheel running and gene therapy alone, but their combination was most beneficial. All treated groups increased in vivo plantarflexor torque over both mdx and mdxR mice. mdx and mdxR mice displayed ∼3-fold lower diaphragm force and power compared to WT values. Treated groups demonstrated partial improvements in diaphragm force and power, with mdxRGT2 mice experiencing the greatest improvement at ∼60% of WT values. Evaluation of oxidative red quadriceps fibers revealed the greatest improvements in mitochondrial respiration in mdxRGT1 mice, reaching WT levels. Interestingly, mdxGT2 mice displayed diaphragm mitochondrial respiration values similar to WT but mdxRGT2 animals showed relative decreases compared to the no run group. Collectively, these data demonstrate that either microdystrophin construct combined with voluntary wheel running increased in vivo maximal muscle strength, power, and endurance. However, these data also highlighted important differences between the two microdystrophin constructs. GT1, with the nNOS-binding site, improved more markers of exercise-driven adaptations in metabolic enzyme activity of limb muscles, while GT2, without the nNOS-binding site, demonstrated greater protection of diaphragm strength after chronic voluntary endurance exercise but decreased mitochondrial respiration in the context of running.