To examine the proportion of older adults with diabetes mellitus treated with tight glucose control and the factors associated with this practice. Cross-sectional analysis. Outpatient sites in the Diabetes Collaborative Registry (N=151). Adults aged 75 and older with type 2 diabetes mellitus (N=42,669). Participants were categorized based on glycosylated hemoglobin (HbA1c) and glucose-lowering medications: poor control (HbA1c >9%), moderate control (HbA1c 8-9%), conservative control (HbA1c 7-8%), tight control (HbA1c <7%) with low-risk agents (low risk for hypoglycemia), tight control with high-risk agents, and diet control (HbA1c <7% taking no glucose-lowering medications). We used hierarchical logistic regression to examine participant and site factors associated with tight control and high-risk agents versus conservative or tight control and low-risk agents. Of 30,696 participants without diet-controlled diabetes, 5,596 (18%) had moderate or poor control, 9,227 (30%) had conservative control, 7,893 (26%) had tight control taking low-risk agents, and 7,980 (26%) had tight control taking high-risk agents. Older age, male sex, heart failure, chronic kidney disease, and coronary artery disease were each independently associated with greater odds of tight control with high-risk agents. There were no differences according to practice specialty (endocrinology, primary care, cardiology) in how aggressively participants were managed. One-quarter of U.S. older adults with type 2 diabetes mellitus are tightly controlled with glucose-lowering medications that have a high risk of hypoglycemia. These results suggest potential overtreatment of a substantial proportion of people and should encourage further efforts to translate guidelines to daily practice.