BACKGROUND: First-contact physiotherapists assess and diagnose patients with musculoskeletal disorders, determining the best course of management without prior general practitioner consultation. OBJECTIVES: The primary aim was to determine the clinical and cost-effectiveness of first-contact physiotherapists compared with general practitioner-led models of care. DESIGN: Mixed-method realist evaluation of effectiveness and costs, comprising three main phases: A United Kingdom-wide survey of first contact physiotherapists. Rapid realist review of first contact physiotherapists to determine programme theories. A mixed-method case study evaluation of 46 general practices across the United Kingdom, grouped as three service delivery models: General practitioner: general practitioner-led models of care (no first contact physiotherapists). First-contact physiotherapists standard provision: standard first-contact physiotherapist-led model of care. First-contact physiotherapists with additional qualifications: first-contact physiotherapists with additional qualifications to enable them to inject and/or prescribe. SETTING: United Kingdom general practice. PARTICIPANTS: A total of 46 sites participated in the case study evaluation and 426 patients were recruited; 80 staff and patients were interviewed. MAIN OUTCOME MEASURES: Short Form 36 physical outcome component score and costs of treatment. RESULTS: No statistically significant difference in the primary outcome Short Form 36 physical component score measure at 6-month primary end point between general practitioner-led, first-contact physiotherapist standard provision and first-contact physiotherapist with additional qualifications models of care. A greater number of patients who had first-contact physiotherapist standard provision (72.4%) and first-contact physiotherapist with additional qualifications (66.4%) showed an improvement at 3 months compared with general practitioner-led care (54.7%). No statistically significant differences were found between the study arms in other secondary outcome measures, including the EuroQol-5 Dimensions, five-level version. Some 6.3% of participants were lost to follow-up at 3 months; a further 1.9% were lost to follow-up after 3 months and before 6 months. Service-use analysis data were available for 348 participants (81.7%) at 6 months. Inspecting the entire 6 months of the study, a statistically significant difference in total cost was seen between the three service models, irrespective of whether inpatient costs were included or excluded from the calculation. In both instances, the general practitioner service model was found to be significantly costlier, with a median total cost of £105.50 versus £41.00 for first-contact physiotherapist standard provision and £44.00 for first-contact physiotherapists with additional qualifications. Base-case analysis used band 7 for first-contact physiotherapist groups. A sensitivity analysis was undertaken at band 8a for first-contact physiotherapists with additional qualifications; the general practitioner-led model of care remained significantly costlier. Qualitative investigation highlighted key issues to support implementation: understanding role remit, integrating and supporting staff including full information technology access and extended appointment times. LIMITATIONS: Services were significantly impacted by COVID-19 treatment restrictions, and recruitment was hampered by additional pressures in primary care. A further limitation was the lack of diversity within the sample. CONCLUSIONS: First-contact physiotherapists and general practitioner models of care are equally clinically effective for people with musculoskeletal disorders. Analysis showed the general practitioner-led model of care is costlier than both the first-contact physiotherapist standard provision and first-contact physiotherapist with additional qualifications models. Implementation is supported by raising awareness of the first-contact physiotherapist role, retention of extended appointment times, and employment models that provide first-contact physiotherapists with professional support. FUTURE RESEARCH: Determining whether shifting workforce impacts physiotherapy provision and outcomes across the musculoskeletal pathway. STUDY REGISTRATION: The study is registered as Research Registry UIN researchregistry5033. FUNDING: This award was funded by the National Institute for Health and Care Research (NIHR) Health and Social Care Delivery Research programme (NIHR award ref: 16/116/03) and is published in full in Health and Social Care Delivery Research; Vol. 12, No. 49. See the NIHR Funding and Awards website for further award information.