Background: Proximal humeral fractures, which occur mainly in older adults, account for approximately 4 to 5% of all fractures. Approximately 40% of these fractures are displaced fractures involving the surgical neck. Management of this group of fractures is often challenging and the outcome is frequently unsatisfactory. In particular it is not clear whether surgery gives better outcomes than non-surgical management. Currently there is much variation in the use of surgery and a lack of good quality evidence to inform this decision. Methods/Design: We aim to undertake a pragmatic UK-based multi-centre randomised controlled trial evaluating the effectiveness and cost-effectiveness of surgical versus standard non-surgical treatment for adults with an acute closed displaced fracture of the proximal humerus with involvement of the surgical neck. The choice of surgical intervention is left to the surgeon, who must use techniques that they are fully experienced with. This will avoid 'learning curve' problems. We will promote good standards of non-surgical care, similarly insisting on care-provider competence, and emphasize the need for comparable provision of rehabilitation for both groups of patients. We aim to recruit 250 patients from a minimum of 18 NHS trauma centres throughout the UK. These patients will be followed-up for 2 years. The primary outcome is the Oxford Shoulder Score, which will be collected via questionnaires completed by the trial participants at 6, 12 and 24 months. This is a 12-item condition-specific questionnaire providing a total score based on the person's subjective assessment of pain and activities of daily living impairment. We will also collect data for other outcomes, including general health measures and complications, and for an economic evaluation. Additionally, we plan a systematic collection of reasons for non-inclusion of eligible patients who were not recruited into the trial, and their baseline characteristics, treatment preferences and intended treatment. Discussion: This article presents the protocol for a multi-centre randomised controlled trial. It gives extensive details of, and the basis for, the chosen methods, and describes the key measures taken to avoid bias and to ensure validity.