To provide the most current assessment of real-world healthcare resource utilization (HRU) and costs among patients with non-valvular atrial fibrillation (NVAF) who newly initiated rivaroxaban and apixaban using a large US database.A retrospective weighted cohort design was used with healthcare insurance claims from the Optum Clinformatics Data Mart databases (January 2012-December 2018). The index date was defined as the first dispensing of rivaroxaban or apixaban. Adult NVAF patients with an index date on or after 1 January 2016, ≥ 12 months of continuous eligibility before the index date and ≥ 1 month after, and without prior use of oral anticoagulant were included. The observation period spanned from the index date to the earliest of the end of data availability, end of insurance coverage, or death. Inverse probability of treatment weighting (IPTW) was used to adjust for differences in baseline characteristics between cohorts. All-cause healthcare resource utilization (HRU), including hospitalization, emergency room, and outpatient visits, and healthcare costs, including medical and pharmacy costs, were evaluated from the payer's perspective during the observation period up to 18 and 24 months, separately.In total, 23,822 rivaroxaban and 53,666 apixaban users were included. After weighting, all baseline characteristics were well balanced between cohorts (mean age: 73.8 years, female: 46.6% in both cohorts). Up to 18 months of follow-up, rivaroxaban users incurred significantly lower total healthcare costs compared to apixaban users (cost difference = -$1,121;In this large retrospective analysis, patients with NVAF initiated on rivaroxaban incurred significantly lower healthcare costs compared to those initiated on apixaban, which were primarily driven by significantly lower outpatient visits and costs during the 18- and 24-month follow-up periods.