Background Although only one-third of patients with depression fully respond to antidepressants, there is little evidence to support the optimal next step of treatment for patients with refractory symptoms. Cognitive behavioral therapy (CBT) may be used as an adjunct to usual care (including medication) for primary care patients with refractory depression, and the CoBalT trial was designed to examine the effectiveness of adjunctive use of CBT compared with usual care alone. Methods This double-parallel group, multicentre randomised controlled trial recruited 469 patients aged 18-75 years with refractory depression (on antidepressants for ≥6 weeks, with a Bayesian Depression Inventory [BDI] score ≥14, and meeting International Classification of Diseases [ICD]-10 criteria for depression) from 73 general practices in the UK. We randomized subjects into two groups (stratified by center and minimized by BDI score at baseline, availability of a counselor in the general practice, previous antidepressant medication, and duration of current depressive episode) using computer-generated codes: the usual care group or the usual care plus CBT group, and were followed for up to 12 months. Due to the characteristics of this intervention, it was not possible to conceal the treatment subgroups from subjects, general practice staff, CBT therapists, and research staff. Analysis was performed by intention-to-treat. The primary prognosis was remission, defined as at least a 50% reduction in depressive symptoms (BDI score) at 6 months compared to the baseline period. This trial is registered under ISRCTN38231611. Results Between November 4, 2008 and September 30, 2010, we assigned 235 patients to the usual care group and 234 patients to the CBT plus usual care group. 422 participants (90%) were followed up at month 6 and 396 (84%) were followed up at 12 months. The trial ended on October 31, 2011. Ninety-five subjects (46%) in the intervention group met the criteria for remission at 6 months, compared with only 46 (22%) in the usual care group Interpretation of results Prior to this study, there was no evidence from large randomized controlled trials of the effectiveness of adding CBT as a next step of treatment to depression medication in depressed patients who do not respond to medication. Our study provides strong evidence that CBT as an adjunct to usual care (including antidepressants) is effective in reducing depressive symptoms in the above population.