Early diagnosis of Alzheimer’s disease and Alzheimer’s disease-related dementias (AD/ADRD) through detection of cognitive impairment (CI), especially at the mild stage, would support timely clinical intervention, recruitment of patients into clinical trials earlier in the disease trajectory and provide important information to initiate and guide care planning conversations with patients and their family members. One barrier to effective early identification of AD/ADRD is the lack of a brief, reliable, and validated screening tool that can be patient administered and is easily integrated into the EHR. Brief screeners for CI present an opportunity to increase early detection, but within the busy primary care setting, a screening tool must be fully EHR-integrated and preferably in a form that can be patient-administered (e.g., through a patient portal). We propose to validate an existing assessment of CI that would support universal screenings for AD/ADRD in older adults during primary care visits using an EHR-integrated brief screener in which patients self-report their cognitive abilities. The use of a patient-reported outcome (PRO) for the specific purpose of early detection of AD/ADRD is ideal in being both patient-centered and minimally burdensome while also providing useful ongoing data points that can alert the provider to subtle changes in cognitive function that may indicate a need for follow-up. The Patient Reported Outcome Measurement Information System (PROMIS) Cognitive Function (PCF) item bank contains 63 items that assess cognitive function. The PROMIS item banking approach has the benefits of being flexible, efficient, precise, and highly suitable for integration into electronic administration platforms, such as EHRs. The PCF items are accessible to those with low literacy, available in several languages, and have been tested for bias according to multiple demographic characteristics including race/ethnicity and age. These properties make the PCF item bank an ideal source for development of a screening tool to detect AD/ADRD among diverse populations in clinical settings. In the course of this five-year project, RAND and its partners (UPMC and NU) will conduct both qualitative and secondary data analysis to establish the validity of scores from the PCF item bank; use the resulting data to derive meaningful scoring criteria and interpretation guidance suitable for use in clinical practice; finalize the screener content; and work with at least two large regional health systems to implement the screener in their EHR and evaluate the feasibility and impact of having these data available during the clinical encounter via the EHR on primary care providers’ behaviors.