INTRODUCTION: It remains unclear whether offering psychiatric patients their preferred treatment influences outcomes at the symptom level. OBJECTIVE: To assess whether offering patients with panic disorder with/without agoraphobia (PD/A) a choice between 2 psychotherapies yields superior outcomes to random assignment. METHODS: In a doubly randomised, controlled preference trial (DRCPT), 221 adults with PD/A were randomly assigned to: choosing panic-focused psychodynamic therapy (PFPP) or panic control treatment (PCT; a form of cognitive behavioural therapy); random assignment to PFPP or PCT; or waiting list control. Primary outcomes were PD/A severity, work status and work absences at post-treatment assessment. Outcomes at post-treatment assessment, 6-, 12-, and 24-month follow-ups were assessed using segmented multilevel linear growth models. RESULTS: At post-treatment assessment, the choice and random conditions were superior to the control for panic severity but not work status/absences. The choice and random conditions did not differ during treatment or follow-up for the primary outcomes. For panic severity, PCT was superior to PFPP during treatment (standardised mean difference, SMD, -0.64; 95% confidence interval, CI, -1.02 to -0.25); PFPP was superior to PCT during follow-up (SMD 0.62; 95% CI 0.27-0.98). There was no allocation by treatment type interaction (SMD -0.57; 95% CI -1.31 to 0.17). CONCLUSIONS: Previous studies have found that offering patients their preferred treatment yields small to moderate effects but have not employed designs that could rigorously test preference effects. In this first DRCPT of 2 evidence-based psychotherapies, allowing patients with PD/A to choose their preferred treatment was not associated with improved outcomes. Further DRCPTs are needed.