Background: Nonresponsiveness to therapy is generally acknowledged, but only a few studies have tested switching to psychotherapy. This study is one of the first to examine the malleability of treatment-resistant patients using acceptance and commitment therapy (ACT). Methods: This was a randomized controlled trial that included 43 patients diagnosed with primary panic disorder and/or agoraphobia (PD/A) with prior unsuccessful state-of-the-art treatment (mean number of previous sessions = 42.2). Patients were treated with an ACT manual administered by novice therapists and followed up for 6 months. They were randomized to immediate treatment (n = 33) or a 4-week waiting list (n = 10) with delayed treatment (n = 8). Treatment consisted of eight sessions, implemented twice weekly over 4 weeks. Primary outcomes were measured with the Panic and Agoraphobia Scale (PAS), the Clinical Global Impression (CGI), and the Mobility Inventory (MI). Results: At post-treatment, patients who received ACT reported significantly more improvements on the PAS and CGI (d = 0.72 and 0.89, respectively) than those who were on the waiting list, while improvement on the MI (d = 0.50) was nearly significant. Secondary outcomes were consistent with ACT theory. Follow-up assessments indicated a stable and continued improvement after treatment. The dropout rate was low (9%). Conclusions: Despite a clinically challenging sample and brief treatment administered by novice therapists, patients who received ACT reported significantly greater changes in functioning and symptomatology than those on the waiting list, with medium-to-large effect sizes that were maintained for at least 6 months. These proof-of-principle data suggest that ACT is a viable treatment option for treatment-resistant PD/A patients. Further work on switching to psychotherapy for nonresponders is clearly needed.