Adeno-associated virus vectors are capable of long-term gene transfer without obvious adverse effects in a number of animal models. Over the last two decades, preclinical and early phase clinical trials in cystic fibrosis and alpha-1 antitrypsin deficiency were undertaken to test the feasibility of this approach. The results of those studies have been important since they have indicated that in vivo gene transfer is feasible and relatively safe. In addition, a number of key limitations to the current generation of AAV2 gene therapy vectors have been defined. The information about these limitations has been used to develop newer AAV vector approaches, based on new mutant and alternative serotype capsids and enhanced promoter systems. The evaluation of safety and efficacy of these newer agents is ongoing.