Continuous infusion ketamine has demonstrated efficacy in the management of refractory seizures yet does not suppress respiratory drive like other continuous infusion anesthetics (CIAs) used for this indication. The aim of this study is to describe the safety and efficacy of ketamine infusions in patients with refractory seizure without intubation. We analyzed a retrospective cohort of adult patients who were not intubated at the time of initiation of a ketamine infusion for refractory seizures. Seizure onset was determined to be the first clinical seizure or nonconvulsive seizure reported on electroencephalography (EEG) or the start time of EEG for patients with seizures at EEG initiation. Responders were defined as patients with resolution of seizures within 24h of initiation without the need for intubation or an additional CIA. A total of 28 patients were analyzed (median age 62years, 64.3% female, 82.1% African American, 82.1% with structural seizure etiology, and 85.7% focal seizures). Of patients, 78.5% received an initial bolus averaging 0.9mg/kg, and the majority (89.3%) were initiated on an infusion at 10μg/kg/minute; median duration was 39.8h. Maximum infusion rates were 10μg/kg/minute in 16 patients, 20μg/kg/minute in seven patients, and 30μg/kg/minute in five patients. Of patients, 71.4% were assessed to be ketamine responders; six of the eight nonresponders required intubation and an additional CIA. Responders were 80% more likely to have received ketamine 5 or more hours earlier than nonresponders. Hypotension (systolic blood pressure < 90mmHg) occurred in 31.8% of patients receiving only ketamine, which did not correlate with duration or maximum dose of ketamine; hypertension (systolic blood pressure > 180mmHg) occurred in 39.3%. Ketamine represents a safe and effective treatment option for refractory seizures and has potential to reduce morbidity associated with intubation in a carefully selected patient population. Early initiation may increase the likelihood of success.