Aspergillus and cytomegalovirus are major causes of morbidity/mortality after haploidentical hematopoietic transplantation. The high degree of mismatching makes cell immunotherapy impossible as it would result in lethal graft-versus-host disease (GvHD). We generated large numbers of donor T-cell clones specific for Aspergillus or cytomegalovirus antigens. We identified clones potentially responsible for causing GvHD by screening them for cross-reactivity against recipient mononuclear cells. Non-recipient reactive, pathogen-specific clones were infused soon after transplantation. They were CD4+ and produced high levels of interferon-gamma and low levels of interleukin-10. In 46 control transplant recipients who did not receive adoptive therapy, spontaneous pathogen-specific T cells occurred in low frequency 9 to 12 months after transplantation and displayed a non-protective low interferon-gamma/high interleukin-10 production phenotype. In the 35 recipients who received adoptive therapy, one single infusion of donor alloantigen-deleted, pathogen-specific clones in the dose range of 10(5) to 10(6) cells/kg body weight did not cause GvHD and induced high-frequency T-cell responses to pathogens, which exhibited a protective high interferon-gamma/low interleukin-10 production phenotype within 3 weeks of infusion. Frequencies of pathogen-specific T cells remained stable over time, and were associated with control of Aspergillus and cytomegalovirus antigenemia and infectious mortality. This study opens new perspectives for reducing infectious mortality after haploidentical transplantations.