Acceptance and commitment therapy (ACT) has been shown to have broad applicability to different diagnostic groups, and there are theoretical reasons to consider its use with clients with chronic mental health problems. We report an innovative treatment development evaluation of ACT for a heterogeneous group of "treatment-resistant clients" (N = 10) who had attended a mean of 3.5 previous psychological interventions. All clients had Axis I presentations and half met diagnostic criteria for Axis II disorders. Functioning, assessed at pre- and postintervention, and at 6- and 12-month follow-up, showed improvements over time on all primary outcome measures, driven largely by significant changes occurring between baseline and 6-month follow-up. Improvements were associated with ACT processes of change. The data thus suggest that a broad range of clients who had not benefited from standard care may benefit from ACT. (Contains 2 tables, 1 figure and 3 videos.)