Background: Inherited GATA3 variants (rs3824662) have been described in higher frequency in both adults and children with Ph-like acute lymphoblastic leukemia (Ph-like ALL) (Perez-Andreu, Nat Genetics 2013; Jain, ASH 2017). The clinical outcomes of Ph-like ALL are well known, and as we have previously described, it as a high-risk subtype of ALL in both children and adults (Roberts, NEJM 2014; Roberts, JCO 2017; Jain, Blood 2017). However, the clinical outcomes of ALL patients with different germline variants of GATA3 that are commonly seen in Ph-like ALL is unknown. Methods: Of the newly diagnosed patients treated at MD Anderson Cancer Center (MDACC) with B-ALL, we performed analyses for the GATA3 rs3824662 variant in 85 patients (42 Ph-like ALL, 43 non-Ph-like ALL). Testing for Ph-like ALL, CRLF2 rearrangement, and GATA3 genotypes was performed as previously described (Jain, ASH 2017). Results: Of the 85 patients, the rs3824662 AA genotype was identified in 22 patients, the AC genotype in 33 patients, and the CC genotype in 29 patients. There was a trend towards increased frequency of Hispanic heritage in the AA genotype (86%) vs AC genotype (61%), vs the CC genotype (31%). Additionally, amongst patients with the AA genotype, 91% had the Ph-like ALL vs 45% of the AC genotype vs 10% of the CC genotype. The median white blood count at diagnosis was 52.1 x 10⁹/L in the AA genotype group, 20 x 10⁹/L in the AC genotype group, and 5.9 x 10⁹/L in the CC genotype group. The 5 year event-free-survival (EFS) of the CC group was 55% vs 36% of the AC group vs 14% of the AA group. The EFS between the 3 genotypic groups was statistically significant with a p value of 0.018. The 5 year overall survival (OS) of the CC group was 64% vs 42% of the AC group vs 14% of the AA group. Conclusions: We observed significant differences in the EFS between patients with 0, 1, 2 copies of this genetic polymorphism in the GATA3 gene. The role of genomic AA GATA3 variant on the biology of ALL and understanding of the downstream mechanisms needs to be determined.