Background: Several randomized trials demonstrated an overall survival (OS) benefit to triplet chemotherapy (5-FU, oxaliplatin, and irinotecan) plus bevacizumab compared to doublet chemotherapy plus bevacizumab in patients with newly diagnosed mCRC, though with higher rates of neutropenic fever and diarrhea. These trials used high-dose continuous infusion 5-FU (3,200 mg/m2/48 hours), but national guidelines suggest low-dose continuous infusion 5-FU (2,400mg/m2/48 hours) due to lower tolerance in U.S. patients. We performed a retrospective cohort study in a U.S. population to compare the effectiveness of low- vs high-dose continuous infusion 5FU in patients undergoing triplet chemotherapy plus bevacizumab for newly diagnosed mCRC. Methods: We used the nationwide Flatiron Health electronic health record-derived database, comprising de-identified patient-level structured and unstructured data curated via technology-enabled abstraction from ~280 cancer clinics. Patients with newly diagnosed mCRC who initiated triplet chemotherapy plus bevacizumab between 10/23/14 and 10/31/22 with non-missing height, weight, and dosing data were included. 5FU infusion doses of 2,000 – 2,799 mg/m2 were categorized as low-dose while those >/= 2,800 mg/m2 were categorized as high-dose. OS from first-line treatment initiation was assessed by dosing group using the Kaplan Meier method with log-rank testing. Multivariable analysis with adjustment for pre-specified variables with imbalance (defined by standardized difference < - 0.10 or > 0.10) was performed using multivariable Cox proportional hazards modeling. Missing data were imputed using multiple imputation with chained equations. Results: Among 320 patients included, 244 received low-dose and 76 patients received high-dose continuous infusion 5FU. Imbalance between low- and high-dose groups was observed in the use of 5FU bolus (23% vs 3%), year of metastatic diagnosis 2018 - 2022 (80% vs 55%), mean irinotecan dose (156.4 vs 160.5 mg/m2), left-sided primary (73% vs 64%), MMR deficiency (3% vs 6%) and mean CEA (478 vs 691 µg/L). In univariate analysis, median OS was 33 months (95% CI 26 – 39) in the low-dose group compared to 34 months (95% CI 24 – 48) in the high-dose group (HR 1.16; 95% CI 0.79 – 1.69; p = 0.44). After adjusting for imbalanced covariates, there was no difference in hazard of death between low- and high-dose groups (HR 0.99; 95% CI 0.64 – 1.53; p = 0.97). Conclusions: In this real-world, U.S. population with newly diagnosed mCRC receiving triplet chemotherapy plus bevacizumab, there was no association with survival between low- vs high-dose continuous infusion 5FU. This supports current guidelines recommending low-dose continuous infusion 5FU (2,400mg/m2/48 hours) for patients with mCRC receiving triplet chemotherapy in the U.S.