PROJECT SUMMARY Older adults with Alzheimer's Disease (AD) and related dementia disorders are at increased risk of a motor vehicle crash due to impairment of the cognitive and sensory functions necessary for safe driving.1 Because there are few sources of long-term longitudinal data on dementia and motor vehicle crashes, and even fewer that include the prodromal stage of AD related dementia called mild cognitive impairment (MCI), very little is known about how the risk of a crash changes over the course of AD and how a crash may change the clinical course of disease. Here, we propose to use a novel data source of Medicare insurance claims linked to licensing, citation, and crash data for more than 1.5 million older drivers over a 13-year period to evaluate longitudinal changes in the risk of a crash and changes in the trajectory of health, health care utilization, and long-term care assistance following a crash, among older drivers with AD. The lack of information on the longitudinal risks and consequences of a crash among older adults with AD represents a critical gap in the literature for two reasons. First, research suggests that there is substantial individual variation in driving ability by stage of disease. While some individuals demonstrate a significant increase in unsafe driving behaviors early in the course of disease, before they are even diagnosed, others are able to drive safely with MCI through the mild to moderate stage of AD.2 As the number of older drivers with AD grows, providers and organizations that serve older adults are increasingly focused on developing interventions to promote “safe-mobility,” allowing older adults to drive for as long as is safely possible. However, to be effective, interventions need to account for longitudinal changes in the risk of a crash over time, and how these trajectories may vary across individual characteristics. Second, the effects of a crash can have long-lasting physical, psychological and social consequences. Current studies have identified short term effects of a crash,5 but no study has followed crash-involved-drivers with AD for long enough to understand how the impact of a crash manifests over time. Individuals with AD are already at higher risk of comorbidity, hospitalization, and early transition to long term care. Thus, given the rise in licensed drivers with AD, it is critical to understand how a crash may compound these risks. A major barrier to identifying periods of high crash risk and the contributing factors is the lack of longitudinal data on motor vehicle crashes. Instead, most studies have evaluated individuals retrospectively at a particular stage of the disease or followed older adults for relatively short periods of time and without knowledge of their cognitive status, making it difficult to identify longitudinal patterns of crash risk. Here, we propose to use 13 years of Medicare claims data linked to licensing, crash and citations for all drivers with a license in New J...