Background Each year in England and Wales, 150,000 people die from cancer, of whom 110,000 will suffer from cancer pain. Research highlights that cancer pain remains common, severe and undertreated, and may lead to hospital admissions. Objective To develop and evaluate pain self-management interventions for community-based patients with advanced cancer. Design A programme of mixed-methods intervention development work leading to a pragmatic multicentre randomised controlled trial of a multicomponent intervention for pain management compared with usual care, including an assessment of cost-effectiveness. Participants Patients, including those with metastatic solid cancer (histological, cytological or radiological evidence) and/or those receiving anti-cancer therapy with palliative intent, and health professionals involved in the delivery of community-based palliative care. Setting For the randomised controlled trial, patients were recruited from oncology outpatient clinics and were randomly allocated to intervention or control and followed up at home. Interventions The Supported Self-Management intervention comprised an educational component called Tackling Cancer Pain, and an eHealth component for routine pain assessment and monitoring called PainCheck. Main outcome measures The primary outcome was pain severity (measured using the Brief Pain Inventory). The secondary outcomes included pain interference (measured using the Brief Pain Inventory), participants’ pain knowledge and experience, and cost-effectiveness. We estimated costs and health-related quality-of-life outcomes using decision modelling and a separate within-trial economic analysis. We calculated incremental cost-effectiveness ratios per quality-adjusted life-year for the trial period. Results Work package 1 – We found barriers to and variation in the co-ordination of advanced cancer care by oncology and primary care professionals. We identified that the median time between referral to palliative care services and death for 42,758 patients in the UK was 48 days. We identified key components for self-management and developed and tested our Tackling Cancer Pain resource for acceptability. Work package 2 – Patients with advanced cancer and their health professionals recognised the benefits of an electronic system to monitor pain, but had reservations about how such a system might work in practice. We developed and tested a prototype PainCheck system. Work package 3 – We found that strong opioids were prescribed for 48% of patients in the last year of life at a median of 9 weeks before death. We delivered Medicines Use Reviews to patients, in which many medicines-related problems were identified. Work package 4 – A total of 161 oncology outpatients were randomised in our clinical trial, receiving either supported self-management (n = 80) or usual care (n = 81); their median survival from randomisation was 53 weeks. Primary and sensitivity analyses found no significant treatment differences for the primary outcome or for other secondary outcomes of pain severity or health-related quality of life. The literature-based decision modelling indicated that information and feedback interventions similar to the supported self-management intervention could be cost-effective. This model was not used to extrapolate the outcomes of the trial over a longer time horizon because the statistical analysis of the trial data found no difference between the trial arms in terms of the primary outcome measure (pain severity). The within-trial economic evaluation base-case analysis found that supported self-management reduced costs by £587 and yielded marginally higher quality-adjusted life-years (0.0018) than usual care. However, the difference in quality-adjusted life-years between the two trial arms was negligible and this was not in line with the decision model that had been developed. Our process evaluation found low fidelity of the interventions delivered by clinical professionals. Limitations In the randomised controlled trial, the low fidelity of the interventions and the challenge of the study design, which forced the usual-care arm to have earlier access to palliative care services, might explain the lack of observed benefit. Overall, 71% of participants returned outcome data at 6 or 12 weeks and so we used administrative data to estimate costs. Our decision model did not include the negative trial results from our randomised controlled trial and, therefore, may overestimate the likelihood of cost-effectiveness. Conclusions Our programme of research has revealed new insights into how patients with advanced cancer manage their pain and the challenges faced by health professionals in identifying those who need more help. Our clinical trial failed to show an added benefit of our interventions to enhance existing community palliative care support, although both the decision model and the economic evaluation of the trial indicated that supported self-management could result in lower health-care costs. Future work There is a need for further research to (1) understand and facilitate triggers that prompt earlier integration of palliative care and pain management within oncology services; (2) determine the optimal timing of technologies for self-management; and (3) examine prescriber and patient behaviour to achieve the earlier initiation and use of strong opioid treatment. Trial registration Current Controlled Trials ISRCTN18281271. Funding This project was funded by the National Institute for Health Research Programme Grants for Applied Research programme and will be published in full in Programme Grants for Applied Research; Vol. 9, No. 15. See the NIHR Journals Library website for further project information.