Introduction: Diverting patients away from the emergency department (ED) has been proposed as a solution for reducing ED overcrowding. The objective of this systematic review is to examine the effectiveness of diversion strategies designed to either direct patients seeking care at an ED to an alternative source of care. Methods: Seven electronic databases and grey literature were searched. Randomized/controlled clinical trials and cohort studies assessing the effectiveness of pre-hospital and ED-based diversion interventions with a comparator were eligible for inclusion. Two reviewers independently screened the studies for relevance, inclusion, and risk of bias. Intervention effects are reported as proportions (%) or relative risks (RR) with 95% confidence intervals (CI). Methodological and clinical heterogeneity prohibited pooling of study data. Results: From 7,306 citations, ten studies were included. Seven studies evaluated a pre-hospital diversion strategy and three studies evaluated an ED-based diversion strategy. The impact of diversion on subsequent health services was mixed. One study of paramedic practitioners reported increased ED attendance within 7 days (11.9% vs. 9.5%; p=0.049) but no differences in return visits for similar conditions (75.2% vs. 72.1%; p=0.64). The use of paramedic practitioners was associated with an increased risk of subsequent contact with health care services (RR=1.21, 95% CI 1.06, 1.38), while the use of deferred care was associated with no increase in risk of subsequently seeking physician care (RR=1.09, 95% CI 0.23, 5.26). While two studies reported that diverted patients were at significantly reduced risk for hospitalization, two other studies reported no significant differences between diverted or standard care patients. Conclusion: The evidence regarding the impact of pre-hospital and ED-based diversion on ED utilization and subsequent health care utilization is mixed. Additional high-quality comparative effectiveness studies of diversion strategies are required prior to widespread implementation.