Background. Type 2 diabetes mellitus (T2DM) disproportionately impacts low socioeconomic status and ethnic/racial minority groups. Research suggests patient activation (i.e., the knowledge, confidence, and skills to manage one’s health) may be associated with better self- management and diabetes clinical control; however, these associations remain largely understudied among diverse primary care patients. This study aimed to: 1) examine the relationships between patient activation and diabetes clinical control indicators (glycosylated hemoglobin [A1c], low-density lipoprotein cholesterol, and systolic blood pressure); 2) examine the relationships between patient activation and self-management behaviors; 3) determine if self-management behaviors act as indirect mechanisms in the association between patient activation and clinical control; and assess whether the relationships among these variables differed across healthcare systems. Design. This cross-sectional study used data collected from 297 participants who completed the baseline assessment of a cluster randomized pragmatic trial testing a medical-assistant health coaching intervention in adults with poorly controlled T2DM from two demographically distinct healthcare systems. Multiple linear regression models tested the associations between 1) patient activation and indicators of diabetes clinical control, and 2) patient activation and self-management. Path analysis tested the indirect effect of self management in the associations between patient activation and clinical control, and the moderating effect of healthcare system. Results. Patient activation was not significantly associated with either of the three clinical control indicators (all ps > .05). Patient activation was significantly associated with overall self-management (B = 0.16, p < .05), healthful diet (B = 0.02, p < .05), low-fat, produce-rich diet (B = 0.02, p < .05), and physical activity (B = 0.03, p < .05). Patient activation was not significantly associated with blood glucose monitoring or medication adherence. In mediation analyses, patient activation was (unexpectedly) positively related to A1c indirectly through overall self- management, B = 0.01 (95% CI: .00, .01). No other significant indirect effects or evidence for moderation were observed. Conclusion. This study sought to clarify the unique role of patient activation in relation to self-management and clinical control in a diverse primary care sample. The lack of consistent associations among study variables underscores the complexity of achieving optimal T2DM outcomes.