Background: While the understanding of AML has been improved over decades, the outcome of relapsed/refractory AML remains dismal. Venetoclax-based regimen has not yet been systematically assessed in R/R AML. Clinical trials of VEN-based regimen designed for R/R AML are urgently needed to further evaluate its efficacy and safety among these patients. We innovatively combined VEN with azacytidine and cytarabine (termed as VAA). Methods: a) Patients: Eligible patients aged 18-70 years and had a confirmed diagnosis of relapsed/refractory AML. b) Trial design and regimens: Phase 2, open-label, single-arm, multicenter study of VAA regimen in R/R AML patients. A stepwise ramp-up of VEN dosing was adopted to achieve the target dose (400 mg), beginning with 100mg on day 1, and escalated to 200mg on day 2, and 400 mg on day 3-day 9. Cytarabine (10mg/m²) was administered by subcutaneous injection twice daily from day 1 to 10, and AZA (75mg/m²) was administered by intravenous injection once daily from day 1 to 7. c) End points and assessments: Primary efficacy endpoints were CRc including CR and CRi, PR and MLFS per IWG for AML. Secondary endpoints were OS and DOR. The safety endpoints were adverse events graded per CTCAE. d) Statistics analysis: As of the date of submission, 29 patients have completed efficacy evaluation and 1 patient is still in the trial. This study was designed to detect a target CRc rate of 65% (historical rate 40%), with 80% power and type 1 error rate at 0.05, and the planned sample size was 31 patients. All P values were two sided. Results: a) Patients: 30 R/R AML patients were enrolled.b) Efficacy: The CRc rate among all patients was achieved in 62.1% (18/29), including CR in 34.5% and CRi in 27.6% patients, following with 13.8% patients obtaining PR. CRc rates were preserved across subgroups with 59.1% (13/22) in ELN intermediate/adverse group, 62.5% (5/8) in those who ≥65 years and 50.0% (7/14) in those whose baseline bone marrow count ≥30%. Noticeably, CRc rate in patients diagnosed with M4/5 reached 72.7% (8/11). With a median follow-up of 6.95 months, median OS and median DOR are not yet reached. Estimated 12-month OS was 86%, and Estimated 12-month DOR was 62.6%. In ELN intermedia/adverse risk patients, the median DOR was 7.5 months. Estimated 12-month OS rate and DOR rate were 84.3% and 63.8%, respectively. Patients with prior bone marrow count less than 30% and with no prior HMA-based therapy were associated with favorable estimated 12-months DOR rate. The regimen had similar efficacy in patients younger than 65 yeas compared with those >65 years. c) Safety: No TLS events were observed. No treatment-related deaths occurred. Conclusions: VAA regimen represents a feasible and safe new choice for R/R AML patients, especially with satisfactory effect in ELN intermediate/adverse group and in those with advanced ages. Clinical trial information: chictr2100051119.