Background: Salvage chemotherapy regimens for patients with relapsed/refractory acute myeloid leukemia (AML) are associated with complete response rates of 30 ‐ 60%. Determining the superiority of one treatment over another is difficult due to the lack of comparative data. Clofarabine and cladribine based regimens appear to be superior to combinations of mitoxantrone, etoposide, and cytarabine (MEC). However, there are no data comparing treatments with these purine analogs to each other. Therefore, we conducted a retrospective study of GCLAC (clofarabine 25 mg/m2 IV days 1-5, cytarabine 2 gm/m2 IV days 1‐5, and G‐CSF) and CLAG (cladribine 5 mg/m2 IV days 1‐5, cytarabine 2 gm/m2 IV days 1‐5, and G‐CSF). Methods: We identified 41 consecutive patients with pathologically diagnosed relapsed or refractory AML who received either GCLAC or CLAG between 2011 and 2014. The primary outcome was the complete response rate (CRi or CR) as defined by the International Working Group. Secondary outcomes included the percentage of patients who underwent allogenic stem cell transplant, relapse free survival (RFS), and overall survival (OS). Fisher’s exact and Wilcoxon Rank Sum tests were used to compare patient characteristics and response rates. The Kaplan Meier method and Log Rank tests were used to evaluate RFS and OS. Results: We found no significant differences in the baseline characteristics of patients treated with GCLAC (n=22) or CLAG (n=19) including age, race, gender, organ function, or cytogenetic risk group (table 1). There were also no significant differences in the percentage of relapsed patients (36% vs. 21%), the average duration of the previous remission (28.6 vs. 19.4 months) or in their previous therapy. An anthracycline-‐based “7+3” regimen was given to 82% of the GCLAC patients and to 90% of the CLAG patients. The outcomes with these two regimens were also not significantly different. Patients treated with GCLAC had a 64% CR/CRi rate compared with 47% for CLAG patients (p=0.36). 45% GCLAC patients underwent allogeneic stem cell transplant compared with 26% of CLAG patients (p=0.32). The median RFS on GCLAC and CLAG respectively was 1.59 years [0.41, non-estimable (NE)] and 1.03 years [0.49, 1.03], (p=0.75). The median OS was 1.03 years [0.52, NE] and 0.70 years [0.28, 1.11], (p=0.08). Given the similarities of these regimens, we combined the data sets to compare the OS for patients with refractory AML to relapsed AML. The OS for patients with refractory AML was not significantly worse than patients with relapsed AML (0.94 years [0.36, 1.3] vs. 1.11 years [0.46, not evaluable]; p=0.49). Conclusion: We find no significant differences in outcomes using GCLAC or CLAG for relapsed/refractory AML patients. The trends in outcome that favored GCLAC are likely explained by trends in patient populations (e.g. longer first remission for GCLAC patients). Since our results are similar to the published reports describing these regimens, we feel the choice of regimen can be based on other considerations such as cost. We do find the efficacy of both regimens in refractory AML to be encouraging. However, we recognize that overall survival of one year is not acceptable and that most relapsed/refractory patients should be entered into clinical trials.