Background: Many decisions in the emergency department (ED) may benefit from patient involvement, even though this setting has been considered least conducive to shared decision-making (SDM). Objectives: The objective was to conduct a systematic review to evaluate the approaches, methods, and tools used to engage patients or their surrogates in SDM in the ED. Methods: Five electronic databases were searched in conjunction with contacting content experts, reviewing selected bibliographies, and conducting citation searches using the Web of Knowledge database. Two reviewers independently selected eligible studies that addressed patient involvement and engagement in decision-making in the ED setting via the use of decision support interventions (DSIs), defined as decision aids or decision support designed to communicate probabilistic information on the risks and benefits of treatment options to patients as part of an SDM process. Eligible studies described and assessed at least one of the following outcomes: patient knowledge, experiences and perspectives on participating in treatment or management decisions, clinician or patient satisfaction, preference for involvement and/or degree of engagement in decision-making and treatment preferences, and clinical outcomes (e.g., rates of hospital admission/readmission, rates of medical or surgical interventions). Two reviewers extracted data on study characteristics, methodologic quality, and outcomes. The authors also assessed the extent to which SDM interventions adhered to good practice for the presentation of information on outcome probabilities (eight probability items from the International Patient Decision Aid Standards Instrument [IPDASi]) and had comprehensive development processes. Results: Five studies met inclusion criteria and were synthesized using a narrative approach. Each study was of satisfactory methodologic quality and used a DSI to engage patients or their surrogates in decision-making in the ED across four domains: 1) management options for children with small lacerations; 2) options for rehydrating children presenting with vomiting or diarrhea or both; 3) risk of bacteremia (and associated complications), tests, and treatment options for febrile children; and 4) short-term risk of acute coronary syndrome (ACS) in adults with low-risk nontraumatic chest pain. Three studies had poor IPDASi probabilities and development process scores and lacked development informed by theory or involvement of clinicians and patients in development and usability testing. Overall, DSIs were associated with improvements in patients' knowledge and satisfaction with the explanation of their care, preferences for involvement, and engagement in decision-making and demonstrated utility for eliciting patients' preferences and values about management and treatment options. Two computerized DSIs (designed to predict risk of ACS in adults presenting to the ED with chest pain) were shown to reduce health care use without evidence of harm. None of the studies reported lack of feasibility of SDM in the ED. Conclusions: Early investigation of SDM in the ED suggests that patients may benefit from involvement in decision-making and offers no empirical evidence to suggest that SDM is not feasible. Future work is needed to develop and test additional SDM interventions in the ED and to identify contextual barriers and facilitators to implementation in practice.