BACKGROUND: Proximal phalanx finger shaft fractures are common and can impair hand function. There is controversy, but no high-quality evidence, on how they are best treated. We compared the clinical and cost-effectiveness of surgery versus non-surgical splint treatment. OBJECTIVE: The primary objective was to compare hand function following surgical fixation with hand function following non-surgical splint treatment using the Hand Health Profile of the Patient Evaluation Measure at 6 months post randomisation. DESIGN: Pragmatic multicentre, parallel superiority randomised (1 : 1) trial. SETTING: Twenty-four acute hospitals in the United Kingdom National Health Service. PARTICIPANTS: Patients ≥ 16 years with one or more proximal phalanx shaft fracture(s), which can be treated via either surgery or non-surgical splint treatment. Patients with intra-articular, base-metaphyseal, neck, open proximal phalanx fractures, injury ≥ 14 days or unable to adhere to trial procedures/complete questionnaires were excluded. INTERVENTIONS: Surgery was any mode of surgical fixation that was considered as appropriate by the treating specialist. Non-surgical splint treatment consisted of any technique/material used in routine care, which may involve manipulation of the fracture with analgesia or local anaesthetic, and subsequent bracing through an externally applied support, usually performed in a clinic or therapy room environment. MAIN OUTCOME MEASURES: The primary outcome measure was the Hand Health Profile of the Patient Evaluation Measure (possible range 11-77, higher scores indicate worst function). Measurements were collected at 6 weeks, 3, 6 and 12 months; 6 months was the primary outcome time point. The primary health outcome for economic evaluation was quality-adjusted life-years in accordance with National Institute for Health and Care Excellence guidelines. RESULTS: Between 9 November 2020 and 2 February 2023, 113 participants were randomised to surgery (n = 56) or non-surgical treatment (n = 55); 2 were excluded. Participants were 60% male, with mean age of 38 years. Treatment arms were balanced. Fifty-three participants in the surgical and 46 in the non-surgical group were included in the primary analysis. At 6 months, the mean Patient Evaluation Measure was 27.1 (standard deviation = 13.6, n = 48) in the surgical group and 25 (standard deviation 12.4, n = 41) in the non-surgical group, with no clinically important difference between groups (adjusted difference in means for surgery vs. non-surgical groups 3, 95% confidence interval -1.6 to 7.7). There were no differences at 6 weeks and 3 months. There were more complications in the surgery group. Surgery was more expensive, resulting in an incremental cost-effectiveness ratio of £39,686 per quality-adjusted life-year, with a 3.2% probability of being cost-effective at the National Institute for Health and Care Excellence threshold of £20,000. CONCLUSIONS: This study supports the conclusion that surgery does not restore better hand function than non-surgical treatment. Surgery results in additional complications and is more expensive. LIMITATIONS: We were unable to further determine if surgery is worse than splint or if the two treatments are equivalent, because of slow recruitment and the smaller than the planned sample size. FUTURE WORK: Research should be directed towards further comparisons of treatments for common hand fractures. FUNDING: This synopsis presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR127292.