Background: Transmission of latent human cytomegalovirus (HCMV) via organ transplantation with post-transplant viral reactivation is extremely prevalent and results in substantial adverse impact on outcomes. Therapies targeting the latent reservoir within the allograft to mitigate viral transmission would represent a major advance. Here, we delivered an immunotoxin (F49A-FTP) that targets and kills latent HCMV aiming at reducing the HCMV reservoir from donor lungs using ex-vivo lung perfusion (EVLP). Methods: HCMV seropositive human lungs were placed on EVLP alone or EVLP + 1mg/L of F49A-FTP for 6 hours (n = 6, each). CD14+ monocytes isolated from biopsies pre and post EVLP underwent HCMV reactivation assay designed to evaluate viral reactivation capacity. Off-target effects of F49A-FTP were studied evaluating cell death markers of CD34+ and CD14+ cells using flow cytometry. Lung function on EVLP and inflammatory cytokine production were evaluated as safety endpoints. Results: We demonstrate that lungs treated ex-vivo with F49A-FTP had a significant reduction in HCMV reactivation compared to controls, suggesting successful targeting of latent virus (76% median reduction in F49A-FTP vs 15% increase in controls, p = 0.0087). Furthermore, there was comparable cell death rates of the targeted cells between both groups, suggesting no off-target effects. Ex-vivo lung function was stable over 6 hours and no differences in key inflammatory cytokines were observed demonstrating safety of this novel treatment. Conclusions: Ex-vivo F49A-FTP treatment of human lungs targets and kills latent HCMV, markedly attenuating HCMV reactivation. This approach demonstrates the first experiments targeting latent HCMV in a donor organ with promising results towards clinical translation.