Background In current non-human primate models, full-thickness porcine grafts could not achieve long-term survival without using potent systemic immunosuppressors. Moreover, the thickness disparity in xeno-PKP proved to be hard to manipulate and may cause several complications which also could prevent the grafts from long survival. Considering the advantages of Descemet's stripping automated endothelial keratoplasty (DSAEK) derived from its ultrathin graft and lower rejection rate, we hypothesize xeno-DSAEK may overcome these imperfections in xeno-PKP. The aim of this study was to explore the feasibility and efficacy of xeno-DSAEK and to investigate the possibility of long-term survival of porcine DSAEK grafts only using local immunosuppressors. Methods Nine rhesus monkeys were divided into two groups. In the DSAEK group (n = 7), Descemet's membrane stripping was performed to establish the bullous keratopathy model followed by the endothelial graft transplant. In the control group (n = 2), only DM stripping was performed. Betamethasone 3.5 mg was injected subconjunctivally every 10 days for a total of 10 times in the DSAEK group. After the surgery, all grafts were evaluated by slit lamp microscopy for at least 6 months. In addition, anterior segment optical coherence tomography (AS-OCT) and confocal laser microscopy (CLM) were hired to evaluate the graft attachment and survival. To investigate the rejection mechanism for xeno-DSAEK, pathological and immunity examinations, including immunohistochemistry, immunofluorescence, T-cell subgroups, cytokine concentration, and anti-GAL antibodies, were performed and compared between the survived and rejected grafts. Results In the control group, the corneal opacity gradually increased with the development of corneal swelling and neovascularization. In the DSAEK group, five recipients (n = 5/7) re-established their transparencies within 30 days after xeno-DSAEK, and four grafts survived >180 days (>180, >180, >298, >270) only by injecting betamethasone subconjunctivally, whereas two grafts (n = 2/7) rejected in 30 days. The AS-OCT and CLM showed the survived grafts attached firmly to the host beds, and the endothelial cell density was adequate (>2000cells/mm(2)) to maintain the graft function at post-operative month 6. The immunohistochemistry showed predominantly T cells (CD4+ and CD8+) and macrophages infiltrated in the rejected grafts, although a few B cells (CD20+) were also observed. Moreover, immunofluorescence demonstrated donor-specific antibody (IgG) and complement (C3c) were also involved in the immune rejection. However, T-cell subgroups and cytokine concentrations (except for IL-6) showed no significant differences between the rejected and survived grafts. Conclusion Xeno-DSAEK is feasible and effective, and the technique may reduce the complications common to xeno-PKP. More importantly, the porcine DSAEK grafts could survive for a long term without using systemic immunosuppressors and exhibit promising future for clinical practice.