Cystic Fibrosis (CF) is a lethal autosomal recessive genetic disorder caused by mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. CFTR transports anions at the apical surface of epithelial membranes and functions in many areas of the body. However in CF, loss of CFTR function in the lungs is the major source of morbidity and mortality. Replacing the defective CFTR in the lungs through gene therapy has the potential to cure the disease. Recombinant adeno-associated virus (AAV) is an effective gene transfer vector and has been used extensively to deliver genes to cells in culture. A number of clinical trials using AAV have been attempted for a variety of diseases, including CF, albeit with limited success. Poor vector transduction efficiency prevents effective gene therapy. We have previously used a technique to greatly increase the transduction efficiency of AAV in human lung tissues by selecting from a library of AAVs using a directed evolution technique. However, this evolution was performed in cultured cells and did not fully represent the in vivo environment in which the AAV would be used. In 2008, a CF pig model was developed to develop a further understanding of the mechanisms of CF and CFTR function. We hypothesized that we could use directed evolution to select for a vector in vivo using the pig, allowing gene therapy studies to be conducted in a physiologically relevant model of CF. We selected a novel AAV variant, called AAV2H22, which is closely related to AAV2 but with greatly increased transduction efficiency in pig airway epithelia. AAV2H22 displayed specific tropism for pig airway epithelia and saturated cell surface receptors, indicating specific binding in those cells. We found that AAV2H22-mediated gene transfer corrected chloride and bicarbonate transport defects both in vitro and in vivo. Importantly, bicarbonate transport was sufficient to normalize pH in the airway surface liquid, resulting in increased bacterial killing likely due to increased activity of antimicrobial peptides. To investigate the mechanics of the increased transduction of AAV2H22, capsid mutants were assayed for transduction efficiency. Two of the five amino acid differences between AAV2 and AAV2H22 lie at the surface and are predicted to alter capsid binding. This is consistent with the results showing specific binding in cultured airway epithelia. This research has important implications for gene therapy and investigations using AAV2H22 will increase our understanding of the biology needed to successfully treat CF.