Background: Seeking to address the well-documented burden of caring for a loved one with a high-grade glioma, this pilot RCT compared the feasibility and preliminary efficacy of a dyadic versus individual yoga intervention as a supportive care strategy for family caregivers. Methods: Caregivers completed self-reported assessments of QOL (SF-36) and caregiver burden (CRA) at baseline. Then, patients and caregivers were randomized to either a dyadic yoga (DY) arm in which caregivers attended the sessions together with the patient, a caregiver yoga (CY) arm in which the caregiver attended all sessions individually or a usual care (UC) arm. Both yoga programs were delivered over 15 sessions either in-person or via videoconference (Zoom) concurrently while patients were undergoing standard radiotherapy. All arms were re-assessed at the end of treatment and then again 3-months later. Results: We consented 76 patient-caregiver dyads (59%) and randomized 23 dyads to the DY, 22 to the CY, and 22 to the UC arm. Attendance in the DY was higher than in the CY group (session means, DY = 12.23, CY = 9.00; P =.06, d =.57). Caregivers (80% female; 78% non-Hispanic White; mean age = 53 yrs) reported significantly more overall subjective benefit in the CY compared to the DY arm (means, CY = 1.39, DY = 1.81, P <.05, d = 1.45). Caregivers slightly favored the in-person delivery over the Zoom-based sessions (means, in person: 1.48; Zoom: 1.82, P =.10; d =.77). A clinically significant, medium effect size was found for improved QOL in favor of the CY over the DY group (means, CY = 49.45, DY = 44.45; F = 3.58, P =.07; d =.67). Caregivers in the CY group also reported less caregiving-related health declines compared to the DY group (means, CY = 2.18; DY = 2.48; F = 4.23, P <.05; d =.42). Caregivers in the CY group reported less caregiving-related financial burden than the UC group (means: CY = 2.79; UC = 3.21; F = 3.32, P =.08; d =.35). Conclusions: Despite lower attendance, caregivers in the CY arm reported greater subjective overall benefit, experienced better mental QOL and less caregiver burden compared with those in the DY and UC comparison arms. It appears that individual rather than dyadic delivery may be a superior supportive care approach for this vulnerable caregiver population. An adequately powered, larger efficacy trial of this intervention strategy is warranted. Clinical trial information: NCT02481349