INTRODUCTION: The contribution of patient support to psilocybin's antidepressant effects remains uncertain. METHODS: Relationships between therapeutic alliance (Scale to Assess Therapeutic Relationship-Patient version; STAR-P), psychedelic experience (Five-Dimensional Altered States of Consciousness Questionnaire and Emotional Breakthrough Inventory; 5D-ASC and EBI) and clinical outcomes (Montgomery-Åsberg Depression Rating Scale; MADRS) were explored using correlation and path analysis for individuals with treatment-resistant depression receiving 25 mg psilocybin with monitoring and support (N = 79). RESULTS: Change from Baseline to Week 3 MADRS scores showed weaker correlations with pre-dosing therapeutic alliance (-0.178) than with measures of the psychedelic experience: EBI (-0.637), Oceanic Boundlessness (-0.508), and Visual Restructuralization (-0.516). Path analysis showed no nominally significant direct effects of therapeutic alliance on Week 3 MADRS scores, but there were nominally significant effects of therapeutic alliance on psychedelic experience (Oceanic Boundlessness (β = 0.28), Visual Restructuralization (β = 0.27), and Auditory Alterations (β = 0.25)). Only one indirect effect of therapeutic alliance on clinical outcome reached nominal significance (via Visual Restructuralization; β = -0.15). Stronger effects were seen on clinical outcomes for psychedelic experience (EBI (β = -0.59), Oceanic Boundlessness (β = -0.53), Visual Restructuralization (β = -0.54), and Auditory Alterations (β = -0.24)). CONCLUSIONS: The therapeutic alliance appeared to facilitate the psychedelic experience, and these experiences in turn had stronger nominally significant direct effects on clinical outcomes. The effects of the alliance itself on therapeutic efficacy were either limited or absent. TRIAL REGISTRATION: EudraCT number: 2017-003288-36; Clinicaltrials.gov identifier: NCT03775200.