Context: Music therapy is a common discretionary service offered within hospice; however, there are critical gaps in understanding the effects of music therapy on hospice quality indicators, such as family satisfaction with care. Objectives: The purpose of this study was to examine whether music therapy affected family perception of patients' symptoms and family satisfaction with hospice care. Methods: A retrospective, cross sectional analysis of electronic medical records from 10,534 cancer patients cared for between 2006-2010 by a large national hospice. Logistic regression was used to estimate the effect of music therapy using propensity scores to adjust for non-random assignment. Results: Overall, those receiving music therapy had higher odds of being female, having longer length of stay, and receiving more services other than music therapy, and lower odds of being married/partnered or receiving home care. Family satisfaction data were available for 1495 (14%) and were more likely available if the patient received music therapy (16% vs 12%, p<.01). There were no differences in patient pain, anxiety, or overall satisfaction with care between those receiving music therapy vs not. Patients who received music therapy were more likely to report discussions about spirituality (OR=1.59, p=.01), marginally less trouble breathing (OR = 0.77, p=.06), and marginally more likely to receive the right amount of spiritual support (OR = 1.59, p=.06). Conclusions: Music therapy was associated with perceptions of meaningful spiritual support and less trouble breathing. The results provide preliminary data for a prospective trial to optimize MT interventions for integration into clinical practice.