OBJECTIVE: To evaluate the efficacy and safety of vigabatrin in pediatric epilepsy. METHODS: We retrospectively reviewed patients with epilepsy treated with vigabatrin over a 2-year period at a pediatric tertiary center. We assessed the relationship between seizure frequency, etiology, vigabatrin dose, adverse events, medication discontinuation reasons, and electroencephalography (EEG) characteristics. RESULTS: One hundred three patients followed at Boston Children's Hospital were treated with vigabatrin and had complete medical records. Within the follow-up interval, 69 (67%) of 103 patients had discontinued vigabatrin therapy. Two patients (1.9%) died during therapy for unknown reasons. Median age at vigabatrin initiation was 8 months (interquartile range [IQR] 5-15). Median starting dose was 48.1 mg/kg per day (IQR 29.8-52.3) with a median target of 100 mg/kg (IQR 81.9-107.9). Median treatment duration was 12.1 months (n = 89, IQR 5.0-22.9) overall, and 13.3 months (IQR 5.2-23.2) for patients who discontinued vigabatrin. The most common reasons for discontinuation were controlled seizures in 31 (43.7%) of 71 and unsatisfactory therapeutic effect in 23 (32.4%) of 71. Median percent seizure reduction from baseline to first follow-up was 83.3% (IQR 27.4-99.8) and 96.7% (IQR 43.3-100) to last follow-up. Twenty-four (38.7%) of 62 patients with a follow-up posttreatment remained seizure-free. Four patients who had initially achieved seizure freedom relapsed. Patients with structural/metabolic etiology had greater median percent seizure reduction at first follow-up than patients with genetic etiology (98.7% vs. 61.4%, respectively, p = 0.001). Hypsarrhythmia resolved after therapy in 18 of 20 (90%, 95% confidence interval [CI] 70-97) patients with pretreatment hypsarrhythmia, and 2 patients presented with hypsarrhythmia posttreatment. Risk of having hypsarrhythmia was reduced by 32% (95% CI 14.9-49.1) posttreatment. SIGNIFICANCE: Vigabatrin is efficacious in all seizure types and resolved hypsarrhythmia in most patients. In this series with a median treatment duration of 12.1 months, vigabatrin had a good safety profile with a low rate of discontinuation due to nonophthalmologic and ophthalmologic adverse effects.