Stroke is the largest cause of adult disability in the United Kingdom (UK), and accounts for 5% of total NHS costs. Early rehabilitation relevant to patients' needs can improve independence and the intensity of activity impacts on the degree of recovery, yet national targets to increase therapy intensity and frequency are not being met. In acute stroke units staff, patients and carers can be more involved in increasing supervised and independent therapeutic activity, which has the potential to expedite discharge and decrease dependency on health and social care services in the longer term. National stroke audits have highlighted non-adherence to therapy intensity and frequency recommendations, but cannot help identify solutions. Our research questions focus on the feasibility and impact of patients, carers and clinicians co-producing and implementing interventions to increase supervised and independent therapeutic patient activity in acute stroke units. The research will be undertaken in three phases. In phase 1 we will complete a rapid evidence synthesis to update the evidence on the efficacy and effectiveness of co-production approaches in acute healthcare settings. In phase 2 we will use Experienced Based Co-Design (EBCD) as our co-production approach to design interventions and evaluate the impact in 2 stroke units, in phase 3 we will implement and evaluate the co-produced interventions developed in phase 3 in a further 2 stroke units. To understand factors influencing development, implementation and embedding of interventions, the Normalisation Process Theory (NPT) approach will be utilised A mixed method, case comparison evaluation will be used. We will collect baseline individual patient data on patient reported outcomes of experiences, wellbeing and satisfaction (PROM/PREM) and utilize Stroke Sentinel National Audit Programme data (SSNAP) data to summarise unit performance against national standards. We will record level of stroke severity of patients receiving care in pre and post intervention periods. Non participant observations will be undertaken to examine social, professional and organisational practices and behavioural mapping methods will be used to record type and number of therapeutic activities in each stroke unit in the pre and post intervention periods. In phase 2, using EBCD methods we will conduct interviews with staff, patients and carers, to produce a 'trigger' film of their narratives. These data will inform the co-production of interventions by patients, carers and staff to increase supervised and independent therapeutic activity. Post intervention, a review of phase 2 implementation data including observational, behavioural mapping and interview data will inform the decision to proceed to phase 3. We will then implement and evaluate the interventions developed in phase 2 in 2 further stroke units. The same pre and post-test intervention measures will be used to evaluate the impact of implementation. To assess how these teams understood and embedded interventions developed and tested elsewhere, NPT methods will be utlised. Anticipated benefits of the study include establishing whether EBCD can be effectively employed in acute healthcare and identifying interventions to increase therapeutic inpatient activity which can be deployed in all stroke units nationally and potentially in other settings including elderly care and trauma and orthopaedic wards.