Under the Medicare Access and Children's Health Insurance Program Reauthorization Act (2015), the Center for Medicare & Medicaid Services (CMS) established the Merit-Based Incentive Payment System Program (MIPS). The program ended the long-standing system that reimbursed clinicians on a fee-for-service basis and shifted Medicare to a system that rewards higher-quality and lower-cost care. Under MIPS, clinicians are required to report data on performance in four domains: quality of care, participation in improvement activities, meaningful use of electronic health records, and cost. CMS weighs each domain to calculate a Composite Performance Score. The program became effective in 2017, with performance scores publicly reported. Starting in 2019, clinicians began receiving payment adjustments of up to 4% of their Medicare Part B payments based on the performance scores achieved in 2017. The adjustment can be upward or downward, and the maximum adjustment is set to increase to 5% in 2020 and 9% from 2022 onward. Four years since MIPS took effect, there is little evidence for its impact on the intended targets (quality and cost of care). Furthermore, similar to many earlier value-based programs that CMS introduced, MIPS has raised concerns about incentivizing strategic responses with little impact on quality of care. For example, to avoid payment reductions, MIPS clinicians may strategically lower their patient risk profiles by referring out high-risk patients, potentially leading to disparities in health outcomes. No research to date has examined the potential unintended effect of MIPS. To fill these knowledge gaps, we propose to examine both desired and potential unintended adverse consequences of MIPS using data from the first three years of the program (2017-2019). This study will use a nationally-representative random sample of MIPS-participating clinicians and their patients from 50 states and the District of Columbia. We will merge several provider-level CMS public reporting data with patient-level Medicare claims data to link participating providers with their attributed patients. Our specific aims are to (1) examine if there are systematic differences in the characteristics of providers and patient profiles by performance scores and (2) evaluate the extent to which performance improvements are associated with changes in healthcare outcomes and patient profiles. This will be the first study to examine the intended (improved healthcare outcome) and unintended effects (changes in patient risk profile) of MIPS. Findings from the first three years of MIPS will inform CMS on the evolution and refinement of the program and serve as the basis for future research on MIPS performance at a larger scale and over later years.