Duchenne muscular dystrophy (DMD) is a common lethal disease for which no effective treatment is available. The lethal consequences of DMD are caused by absence of a structural protein, called dystrophin, from skeletal and cardiac muscle cells. The usefulness of gene replacement as therapy for this disease has been established in transgenic mouse models. Unfortunately, progress toward therapy for human patients has been limited by the characteristics of currently available viral vectors and by lack of a suitable technique for delivery of such vectors to a large mass of muscle cells. Successful gene therapy of DMD will require a vector that can carry most of the dystrophin coding sequence, that can be cheaply produce in large quantities, that can be delivered to a large mass of muscle cells, and that provides stable expression of dystrophin after delivery. We and others have worked to develop such a vector through modification of adenoviruses (Ad). Here we review the characteristics of conventional Ad vectors and new helper-dependent, or gutted, Ad vectors. Gutted Ad vectors contain cis-acting DNA sequences necessary for viral replication and packaging, but are deleted, or gutted, for all viral coding sequences. We found that gutted vectors efficiently delivered full-length dystrophin to the skeletal muscles of dystrophic (mdx) mice. Dystrophic muscles injected with these vectors expressed dystrophin for at least four months post-injection, which was the longest time point tested. These data suggest that gutted vectors will allow delivery and long-term expression of dystrophin.