The US Preventive Services Task Force (USPSTF) recommends aspirin for primary prevention of atherosclerotic vascular disease (ASCVD) when the ASCVD benefit outweighs the risk of gastrointestinal hemorrhage. The complexity and time required to assess aspirin risks and benefits can result in overuse and underuse of aspirin. As part of an NIH-funded study to lower ASCVD risk, we implemented electronic clinical decision support (CDS) algorithms to guide aspirin use based on USPSTF criteria and major bleeding risks. Baseline data was collected for whether aspirin was algorithmically recommended for all patients at their first eligible primary care encounter in 20 clinics over 2012-2014. The analysis excluded patients with CHD and included 6651 adults with diabetes (mean age 55.6, mean 10-year ASCVD risk 27.8%) and 11,682 adults meeting pre-specified criteria for high ASCVD risk without diabetes (mean age 58.4, mean 10-year ASCVD risk 24.7%). Overuse and underuse was determined by comparing concordance with (a) aspirin recommendations and (b) documented aspirin use. The CDS recommended aspirin for 4,139 (63.1%) patients with diabetes and 8,722 (74.7%) without diabetes. Among patients with aspirin recommended, aspirin was not used in 829/4139 (20%) with diabetes and 6493/8722 (74.4%) without diabetes (underuse). Among patients for whom the CDS did not recommend aspirin, aspirin was used in 1448/2969 (59.8%) with diabetes and 1021/2960 (34.4%) without diabetes (overuse). Those with diabetes who were likely to benefit from aspirin use had higher aspirin use rates (less underuse) than similar high CV risk patients without diabetes. However, those with diabetes who were unlikely to benefit from aspirin based on USPSTF criteria and bleeding risks also had higher aspirin use rates (more overuse) than patients without diabetes. Strategies to ensure greater evidence-based use of aspirin, such as providing electronic clinical decision support, may help providers more accurately assess individualized risks and benefits of aspirin.