The in vivo administration of adenovirus vectors frequently elicits a neutralizing antibody response which eliminates or substantially reduces the efficacy of subsequent treatments. Methods to overcome this significant barrier to repeat delivery will be required for the application of adenovirus-based gene therapy in the treatment of chronic disease. We have evaluated the relationship between the initial vector dose and the effectiveness of a second vector administration. C57BL/6 mice injected intravenously with up to 10(7) p.f.u. of a lacZ adenovirus vector, Av1lacZ4, expressed significant levels of human factor IX when injected with 2 x 10(8) p.f.u. of the factor IX vector, Av1H9F, 5 weeks later. An initial dose of 10(8) p.f.u. of Av1lacZ4 completely prevented expression of factor IX following the second administration due to the generation of neutralizing antibody. However, transient immunosuppression with deoxyspergualin (DSG) or cyclophosphamide at the time of initial exposure to 10(8) p.f.u. of Av1lacZ4 prevented the formation of anti-adenovirus neutralizing antibody and permitted an effective second administration of a factor IX vector. Furthermore, transient immunosuppression with cyclophosphamide concomitant with delivery of the factor IX vector enabled an effective administration of a third vector encoding human factor VIII. This approach, together with strategies to prolong the persistence of adenoviral vector expression, should permit long-term therapy with adenovirus-based vectors.