Adeno-associated viruses (AAVs) have been favored for in vivo gene transfer of therapeutic genes because of their low immunogenicity, strong safety record, and high efficiency of transduction of a number of cell types in animal models. 1 However, in a recent gene therapy trial involving delivery of human coagulation factor IX (F. IX) for the treatment of hemophilia, two patients developed a T-cell response to AAV-2 capsid, which was not predicted from preclinical studies in animals. Importantly, transgene expression in the two patients declined to pretreatment levels, suggesting that a cytotoxic T-lymphocyte (CTL) response to the capsid may have been responsible for a loss of transgene-expressing cells. In this issue, Li et al. now provide a detailed analysis of the effect of CTL responses to AAV capsid in mice. 2 The results highlight a problem facing the field: the lack of an appropriate and predictive animal model for immune-mediated loss of hepatocytes after transduction with AAV.