Background: Caregivers maximize health and quality of life of individuals undergoing treatment for cancer by assisting with daily activities, managing complex care, navigating health care systems, and communicating with health care professionals. In metastatic breast cancer (MBC), caregiver roles may be more extensive due to treatment trajectories. Little is known about how caregivers participate in clinical encounters and how their roles in treatment decision-making impact patient engagement. Methods: This mixed-methods analysis explored the effect of the quantity and quality of caregiver participation at clinical encounters in women with MBC. Participation quantity was measured by calculating proportion speaking time (seconds spoken/total appointment time) for oncologists, patients, and caregivers. A summative thematic content analysis using a constant comparative method was utilized to inductively identify caregiver roles during recorded appointments. Caregiver participation quality was then operationalized as the summed number of roles displayed (0-12). Patient participation quality was assessed using the Patient Activation Measure (PAM; scored 0-100 with higher scores indicating more activation) and compared to caregiver quantity and quality of participation using Spearman’s correlations. Results: Fifty-three clinical encounters with verbal caregiver participation were recorded. Patients had a mean age of 58 (SD 10) and were predominately white (72%); demographics of caregivers were not collected. On average, encounters lasted for 30 minutes (SD 12) during which oncologists spoke for 76%, patients for 20%, and caregivers for 4% of the encounter. Identified caregiver roles included: General Support; Management of Treatment or Medication; Treatment History; Decision-Making; Insurance or Money; Pharmacy; Scheduling; Travel Concerns; General Cancer Understanding; Patient Specific Cancer Understanding; Caregiver-Initiated or Emphasis on Symptom Severity; and Caregiver Back-Up of Patient Symptom Description. Caregiver quality scores averaged 5 roles (SD 3) and patients were highly activated (M=65, SD 16). Quality [r(53) = 0.349, p=0.010] and not the quantity [r(53) = 0.213, p=0.126] of caregiver contributions were associated with higher levels of patient activation. Conclusions: Although caregiver time speaking did not impact patient activation, the more ways that the caregiver productively participated in the clinical encounter, the more activated the patient was in their care. Greater integration of the caregiver into treatment decision-making appointments could help to encourage patient engagement when developing a treatment plan for MBC.