Objective. To further conceptualize and operationalize patient activation (PA), using measures from patient, physician, and researcher perspectives. Data Source/Study Setting. Multimethod observation in 2010 within a family medicine clinic. Study Design. Part of an intervention with 130 patients with type 2 diabetes, this observational study further looked at PA in 19 physician-patient dyads. Data Collection. Observations occurred in a teaching hospital, which served as recruiting and study site. Principal Findings. PA correlated with knowledge, self-efficacy, promotion orientation, and exercise intent. Patient-reported PA did not correlate with researcher-observed or physician-reported PA behavior. Researcher-observed PA correlated with physician-observation items. Conclusions. Results provide evidence for measuring different perspectives in studies of PA. When patients report they are activated in self-management, behavior does not indicate they are active in clinical communication, a critical component of collaborative decision making. (Contains 3 tables.)