Background The NHS currently faces increasing demands on accident and emergency departments. Concern has been expressed regarding whether the needs of vulnerable groups are being handled appropriately or whether alternative methods of service delivery may provide more appropriate emergency and urgent care services for particular groups. Objective Our objective was to identify what interventions exist to manage use of the emergency and urgent care system by people from a prespecified list of vulnerable groups. We aimed to describe the characteristics of these interventions and examine service delivery outcomes (for patients and the health service) resulting from these interventions. Review methods We conducted an initial mapping review to assess the quantity and nature of the published research evidence relating to seven vulnerable groups (socioeconomically deprived people and families, migrants, ethnic minority groups, the long-term unemployed/inactive, people with unstable housing situations, people living in rural/isolated areas and people with substance abuse disorders). Databases, including MEDLINE and the Cumulative Index to Nursing and Allied Health Literature, and other sources were searched between 2008 and 2018. Quantitative and qualitative systematic reviews and primary studies of any design were eligible for inclusion. In addition, we searched for UK interventions and initiatives by examining press reports, commissioning plans and casebooks of ‘good practice’. We carried out a detailed intervention analysis, using an adapted version of the TIDieR (Template for Intervention Description and Replication) framework for describing interventions, and an analysis of current NHS practice initiatives. Results We identified nine different types of interventions: care navigators [three studies – moderate GRADE (Grading of Recommendations, Assessment, Development and Evaluations)], care planning (three studies – high), case finding (five studies – moderate), case management (four studies – high), front of accident and emergency general practice/front-door streaming model (one study – low), migrant support programme (one study – low), outreach services and teams (two studies – moderate), rapid access doctor/paramedic/urgent visiting services (one study – low) and urgent care clinics (one systematic review – moderate). Few interventions had been targeted at vulnerable populations; instead, they represented general population interventions or were targeted at frequent attenders (who may or may not be from vulnerable groups). Interventions supported by robust evidence (care navigators, care planning, case finding, case management, outreach services and teams, and urgent care clinics) demonstrated an effect on the general population, rather than specific population effects. Many programmes mixed intervention components (e.g. case finding, case management and care navigators), making it difficult to isolate the effect of any single component. Promising UK initiatives (front of accident and emergency general practice/front-door streaming model, migrant support programmes and rapid access doctor/paramedic/urgent visiting services) lacked rigorous evaluation. Evaluation should therefore focus on the clinical effectiveness and cost-effectiveness of these initiatives. Conclusions The review identified a limited number of intervention types that may be useful in addressing the needs of specific vulnerable populations, with little evidence specifically relating to these groups. The evidence highlights that vulnerable populations encompass different subgroups with potentially differing needs, and also that interventions seem particularly context sensitive. This indicates a need for a greater understanding of potential drivers for varying groups in specific localities. Limitations Resources did not allow exhaustive identification of all UK initiatives; the examples cited are indicative. Future work Research is required to examine how specific vulnerable populations differentially benefit from specific types of alternative service provision. Further exploration, using primary mixed-methods data and potentially realist evaluation, is required to explore what works for whom under what circumstances. Rigorous evaluation of UK initiatives is required, including a specific need for economic evaluations and for studies that incorporate effects on the wider emergency and urgent care system. Funding The National Institute for Health Research Health Services and Delivery Research programme.