BACKGROUND Relapse of primary hematologic disease constitutes an important reason for failure of allogeneic hematopoietic stem cell transplantation (alloHSCT). There are very few treatment modalities for this indications. Therefore, there is a need for novel effective therapies and even more for the prevention of relapse. There are scarce data that azacitidine can be used for these purposes. METHODS At the Polish Adult Leukemia Group, we retrospectively analyzed the results of azacitidine treatment after alloHSCT. Relapsing patients, patients with minimal residual disease/mixed chimerism, and patients in complete remission with high risk of relapse were analyzed separately. There were 17 patients, 6 with myelodysplastic syndrome, 11 with acute myeloid leukemia, 8 male, and overall median age of 56 years (range, 15-78); 7 patients received donor lymphocyte infusion (DLI). RESULTS Patients treated because of relapse received a median of 3 (range, 1-6) cycles of azacitidine, patients receiving preemptive treatment received a median of 4 cycles (range, 2-6), and those on maintenance received a median of 5 cycles (range, 3-5). Toxicity was considerable, especially in relapse-neutropenia (67%), anemia (67%), thrombocytopenia (100%), serious infections (78%)-and preemptive settings. Median overall survival of patients treated for relapse reached 6.8 months (95% confidence interval [CI], 0.7-∞), with better survival observed in patients with temporary disease control (7.7 vs 4.7 mo) and without previous exposure to azacitidine (7.7 vs 3.4 mo). One-year overall survival reached 75% (95% CI, 13%-96%) for preemptive and 50% (95% CI, 0%-91%) for maintenance treatment. DLI did not aggravate graft-versus-host disease. CONCLUSIONS Effectiveness of azacitidine in relapsing patients is disappointing. Azacitidine seems to be promising in preemptive and maintenance settings. Toxicity is considerable. Further research is needed.