CONTEXT: Massage therapy is increasingly used in palliative settings to improve quality of life (QoL) and symptom burden; however, the optimal massage "dosage" remains unclear. OBJECTIVES: To compare three massage dosing strategies among inpatients receiving palliative care consultation. METHODS: At an urban academic hospital, we conducted a three-armed randomized trial examining three different doses of therapist-applied massage to test change in overall QoL and symptoms among hospitalized adult patients receiving palliative care consultation for any indication (Arm I: 10-min massage daily × 3 days; Arm II: 20-min massage daily × 3 days; Arm III: single 20-min massage). Primary outcome measure was single-item McGill QoL question. Secondary outcomes measured pain/symptoms, rating of peacefulness, and satisfaction with intervention. Data were collected at baseline, pre- and post-treatment, and one-day postlast treatment (follow-up). Repeated measure analysis of variance and paired t-test were used to determine significant differences. RESULTS: Total n = 387 patients were 55.7 (±15.49) years old, mostly women (61.2%) and African-American (65.6%). All three arms demonstrated within-group improvement at follow-up for McGill QoL (all P < 0.05). No significant between-group differences were found. Finally, repeated measure analyses demonstrated time to predict immediate improvement in distress (P ≤ 0.003) and pain (P ≤ 0.02) for all study arms; however, only improvement in distress sustained at follow-up measurement in arms with three consecutive daily massages of 10 or 20 minutes. CONCLUSION: Massage therapy in complex patients with advanced illness was beneficial beyond dosage. Findings support session length (10 or 20 minutes) was predictive of short-term improvements while treatment frequency (once or three consecutive days) predicted sustained improvement at follow-up.