Viral infections are agreat risk for successful transplantation. Surprisingly, also the immuneresponse against viruses can affect transplantation outcome, as T cells thatare trained to recognize self-HLA presenting a viral peptide, can make a“mistake” and recognize allogeneic (donor) HLA. This phenomenon is known tooccur frequently, but its clinical relevance remains unknown. In this thesis, we investigated and optimized the methods to detectcross-reactivity of virus-specific T cells in vitro. Illustrating the scope ofthe problem, we found that infection with a single virus induces a broadrepertoire of alloreactive T cells. Conversely, cross-reactivity may also boostanti-viral immunity in immunocompromised patients. Also, we found thatvirus-specific T cells of unrelated individuals can recognize the sameallogeneic HLA much more frequently than anticipated. This knowledge may helpto predict which patient-donor combinations induce alloreactivity.Addressing clinical relevance in transplantation, we showed that the immunepotential against the viral and allogeneic target can be equally strong,depending on the expression of the (allo)epitope. However, clinical relevanceonly truly be determined when the peptides presented by the cross-reactingallo-HLA molecules are identified. A promising allopeptide-defining strategy iscurrently under development at Monash University (Melbourne, Australia),described in the Discussion.