THE success of adenoviral vectors for gene therapy of lung disease in cystic fibrosis (CF) depends on efficient transfer of the complementary DNA encoding the correct version of the cystic fibrosis transmembrane regulator (CFTR) to the affected columnar epithelial cells lining the airways of the lung. Pre-clinical studies in vitro suggest that low doses of adenovirus vectors carrying this CFTR cDNA can correct defective Cl⁻ transport in cultured human CF airway epithelia¹. Here we use mice carrying the disrupted CF gene² to test the efficacy of this transfer system in vivo. We find that even repeated high doses can only partially (50%) correct the CF defect in Cl⁻ transport in vivo and do not correct the Na⁺ transport defect at all. We investigated this discrepancy between the in vivo and in vitro transfer efficiency using CF mouse and human samples, and found that it reflects a difference in the susceptibility to adenovirus-5 transduction of the epithelial cell types dosed in vivo (columnar) and in vitro (basal-cell-like). These studies indicate that more efficient adenoviral gene-transfer vectors and/or refinement of dosing strategies are needed for therapy of CF lung disease.