Purpose: A stepped-wedge cluster-randomised controlled trial was conducted to evaluate the feasibility and effectiveness of a brief psychosocial intervention for depressed cancer patients, delivered by trained front-line health professionals in routine clinical care. Methods: Nine hundred two patients were assessed across four treatment centres which were allocated in random order from control epoch to intervention epoch. Eligible patients had Hospital Anxiety and Depression Scale (HADS) scores of 8 or greater. Of eligible patients, 222 were recruited in control epoch and 247 in intervention epoch. Twenty-seven health professionals (HPs) were trained to deliver the psychosocial intervention consisting of up to four sessions, tailored to patient symptoms and distress. HPs participated in group supervision with a psychiatrist. The primary outcome, analysed by intention to treat, was depression measured with the HADS at 10 weeks after receiving the intervention. Results: At 10-week follow-up, there were no significant differences in HADS score for the 181 patients in control epoch and 177 in intervention epoch (adjusted difference −1.23, 95 % CI −3.81––1.35, p = 0.35). Patients with disease progression who received the intervention experienced significant benefits in unmet practical support needs including care and support, information, and physical and daily living. Conclusion: A brief psychosocial intervention delivered by front-line oncology health professionals is feasible to deliver but is insufficient as a stand-alone treatment for depression in cancer patients. Psychosocial interventions should be targeted to populations most likely to experience benefit.