Objective: Proton pump inhibitors (PPIs) are an effective treatment for eosinophilic esophagitis (EoE); however, only 30% to 60% of patients respond. Common genetic variants in CYP2C19 and STAT6 associate with PPI plasma concentration and magnitude of inflammatory response, respectively. Our objective was to determine if genetic variation in the genes for CYP2C19 and STAT6 influence differentiation between PPI responsive esophageal eosinophilia versus PPI nonresponsive EoE (PPI-REE, PPI-nonresponsive EoE). Methods: Genomic DNA was isolated from 92 esophageal tissue biopsies collected from participants of a prospective clinical trial of high-dose PPI therapy for esophageal eosinophilia in children. Results: Of the 92 patients examined, 57 (62%) were PPI-REE and 35 (38%) were PPI-nonresponsive EoE. Forty-six of the 92 patients were further characterized by pH probe monitoring; there was no association between reflux index and carriage of CYP2C19 Ã 17 (P ¼ 0.35). In children who received a PPI dose between !1.54 and 2.05 mg/kg/day, binary logistic regression modeling showed that carriage of CYP2C19 Ã 17 associated with PPI-nonresponsive EoE (odds ratio (OR) [95% confidence interval (CI)] ¼ 7.71 [1.21, 49.11], P ¼ 0.031). Carriage of STAT6 allelic variant rs1059513 predicts PPI-REE (OR [95% CI] ¼ 6.16 [1.44, 26.4], P ¼ 0.028), whereas carriage of STAT6 rs324011 synergizes with CYP2C19 Ã 17 to predict PPInonresponsive EoE (rs324011 OR [95% CI] ¼ 5.56 [1.33, 20.72], P ¼ 0.022; CYP2C19 Ã 17 OR [95% CI] ¼ 8.19[1.42, 50.57], P ¼ 0.023). Conclusions: Common variants in CYP2C19 and STAT6 associate with a PPI-nonresponsive EoE outcome of PPI therapy for esophageal eosinophilia suggesting that response rates may be improved by adopting a genotypeguided approach to PPI dosing.