Introduction: Agitated patients in the prehospital setting pose challenges for both patient care and emergency medical services (EMS) provider safety. Midazolam is frequently used to control agitation in the emergency department setting; however, limited data exist in the prehospital setting. We describe our experience treating patients with midazolam for behavioral emergencies in a large urban EMS system. We hypothesized that using midazolam for acute agitation leads to improved clinical conditions without causing significant clinical deterioration. Methods: We performed a retrospective review of EMS patient care reports following implementation of a behavioral emergencies protocol in a large urban EMS system from February 2014-June 2016. For acute agitation, paramedics administered midazolam 1 milligram (mg) intravenous (IV), 5 mg intramuscular (IM), or 5 mg intranasal (IN). Results were analyzed using descriptive statistics, Levene's test for assessing variance among study groups, and t-test to evaluate effectiveness based on route. Results: In total, midazolam was administered 294 times to 257 patients. Median age was 30 (interquartile range 24-42) years, and 66.5% were male. Doses administered were 1 mg (7.1%) and 5 mg (92.9%). Routes were IM (52.0%), IN (40.8%), and IV (7.1%). A second dose was administered to 37 patients. In the majority of administrations, midazolam improved the patient's condition (73.5%) with infrequent adverse events (3.4%). There was no significant difference between the effectiveness of IM and IN midazolam (71.0% vs 75.4%; p = 0.24). Conclusion: A midazolam protocol for prehospital agitation was associated with reduced agitation and a low rate of adverse events.