Theoretically, cystic fibrosis transmembrane conductance regulator (CFTR) gene replacement during the neonatal period can decrease morbidity and mortality from cystic fibrosis (CF). In vivo gene transfers have been accomplished in CF patients. Choice of vector, mode of delivery to airways, translocation of genetic information, and sufficient expression level of the normalized CFTR gene are issues that currently are being addressed in the field. The advantages and limitations of viral vectors are a function of the parent virus. Viral vectors used in this setting include adenovirus (Ad) and adeno-associated virus (AAV). Initial studies with Ad vectors resulted in a vector that was efficient for gene transfer with dose-limiting inflammatory effects due to the large amount of viral protein delivered. The next generation of Ad vectors, with more viral coding sequence deletions, has a longer duration of activity and elicits a lesser degree of cell-mediated immunity in mice. A more recent generation of Ad vectors has no viral genes remaining. Despite these changes, the problem of humoral immunity remains with Ad vectors. A variety of strategies such as vector systems requiring single, or widely spaced, administrations, pharmacologic immunosuppression at administration, creation of a stealth vector, modification of immunogenic epitopes, or tolerance induction are being considered to circumvent humoral immunity. AAV vectors have been studied in animal and human models. They do not appear to induce inflammatory changes over a wide range of doses. The level of CFTR messenger RNA expression is difficult to ascertain with AAV vectors since the small size of the vector relative to the CFTR gene leaves no space for vector-specific sequences on which to base assays to distinguish endogenous from vector-expressed messenger RNA. In general, AAV vectors appear to be safe and have superior duration profiles. Cationic liposomes are lipid-DNA complexes. These vectors generally have been less efficient than viral vectors but do not stimulate inflammatory and immunologic responses. Another challenge to the development of clinically feasible gene therapy is delivery mode. Early pulmonary delivery systems relied on the direct instillation of aerosolized vectors, which can result in the induction of adverse reactions because vector is delivered into the lung parenchyma. More recent studies have examined the potential for using spray technologies to target aerosolized AAV vectors to the larger central airways, thereby avoiding alveolar exposure and adverse effects. Comparisons of lung deposition with nebulized delivery of aerosol and spray delivery indicate that spraying results in a more localized deposition pattern (predominantly in the proximal airways) and significantly higher deposition fractions than nebulization. These findings could lead to more efficient and targeted lung delivery of aerosolized gene vectors in the future.