Recombinant adeno-associated virus (rAAV) vectors possess a number of properties that may make them suitable for clinical gene therapy, including being based upon a virus for which there is no known pathology and a natural propensity to persist in human cells. Wild-type adeno-associated viruses (AAVs) are now known to be very diverse and ubiquitous in humans and nonhuman primates, which adds to the degree of confidence one may place in the natural history of AAV, namely that it has never been associated with any human tumors or other acute pathology, other than sporadic reports of having been isolated from spontaneously aborted fetuses. On the basis of this understanding of AAV biology and a wide range of preclinical studies in mice, rabbits, dogs and nonhuman primates, a growing number of clinical trials have been undertaken with this class of vectors. Altogether, over 40 clinical trials have now been approved. Although all previous trials were undertaken using AAV serotype 2 vectors, at least two current trials utilize AAV2 vector genomes cross-packaged or pseudotyped into AAV1 capsids, which appear to mediate more efficient gene delivery to muscle. The explosion of capsid isolates available for use as vectors to over 120 has now provided the potential to broaden the application of AAV-based gene therapy to other cell types.