Background: Previous work has demonstrated treatment (tx) heterogeneity in the care of patients with aGC/GEJ. This study was designed to examine heterogeneity temporal trends and OS in patients with aGC/aGEJ in the United States from 2011 to 2018. Methods: The Flatiron Advanced/metastatic gastric/esophageal cohort electronic medical records (EMR) data were used for this study. Eligible patients were adults receiving anticancer therapy for aGC/GEJ during the study period. Tx patterns were summarized by line of therapy. Drugs were grouped by class. Heterogeneity was measured using the Herfindahl-Hirschman index (HHI); HHI scores range from 0.0000 (complete heterogeneity) to 1.0000 (complete homogeneity). A difference of 0.1000 in HHI scores is considered to be practically meaningful. HHI scores were calculated for each clinic with ≥10 patients. OS was estimated using Kaplan-Meier method. Trend analyses were conducted for HHI scores over time using a linear regression model. Results: There were 2,912 patients who met eligibility criteria. The median age of the study cohort was 67 years; majority were male (70.9%) and white (61.1%). aGC patients comprised the majority (n=1,630, 55.9%). Median OS from the start of first-line (1L) therapy was 12.7 months (95% CI: 12.07, 13.6). The most common 1L regimens were fluoropyrimidine + oxaliplatin (n=651, 22.4%), platinum (ie, cisplatin or carboplatin) + taxane (n=511, 17.5%), and single-agent fluoropyrimidine (n=280, 9.6%). 1,230 patients received second line (2L) and the most common regimens were ramucirumab + taxane (n=203, 16.5%), fluoropyrimidine + oxaliplatin (n=155, 12.6%), and platinum + taxane (n=101, 8.2%). Overall median HHI for 1L was 0.1728 (min/max: 0.0926–0.4380). Median 1L HHI for 2011–2012 was 0.21665 (min/max: 0.1626–0.3156) and was 0.2419 (min/max 0.1716–0.4583) for 2017–2018. There were no significant differences in HHI score over time (P=.078). Overall median HHI for 2L was 0.1309 (min/max: 0.0694–0.2400). Due to small number of sites with ≥10 patients, data in 2L were too limited to conduct time trend analyses. Conclusions: Heterogeneity in 1L treatment of gastric/GEJ cancer persists, despite the continued refinement of guidelines and approval of new drugs in subsequent lines of therapy. Further analyses are needed to examine the relationship between heterogeneity, guideline adherence, and patient outcomes. These future data will shed light on the persistence of heterogeneous treatment patterns observed in the United States.