© 2019 Background: Since 2014, patients with severe treatment-resistant epilepsies (TREs) have been receiving add-on cannabidiol (CBD) in an ongoing, expanded access program (EAP), which closely reflects clinical practice. We conducted an interim analysis of long-term efficacy and tolerability in patients with Lennox-Gastaut syndrome (LGS) or Dravet syndrome (DS) who received CBD treatment through December 2016. Methods: Children and adults with LGS/DS taking stable doses of antiepileptic drugs (AEDs) at baseline were included from 25 EAP sites across the United States. During the 4-week baseline period, parents/caregivers kept diaries of all countable seizure types. Patients received a pharmaceutical formulation of highly purified CBD (Epidiolex®; 100 mg/mL) in oral solution at 2–10 mg/kg/day, titrated until tolerability limit or a maximum dose of 25–50 mg/kg/day. Patient visits were every 2–4 weeks. The percentage change from baseline in median monthly convulsive (ie, major motor) and total seizures was evaluated at 12-week intervals through 96 weeks. The percentages of patients who had ≥50%, ≥75%, and 100% reduction in monthly seizures relative to the baseline period were also evaluated. Adverse events (AEs) were monitored and summarized for the safety analysis set (SAS) through 144 weeks. Results: Of the 607 patients in the SAS, 58 had DS and 94 had LGS (N = 152); 455 patients had other TREs. Twenty-eight percent of LGS/DS patients withdrew, primarily owing to lack of efficacy (20%). LGS/DS patients were taking a median of 3 (0–10) concomitant AEDs. Median treatment duration was 78.3 (range, 4.1–146.4) weeks. Between weeks 12 and 96, median CBD dose ranged from 21 to 25 mg/kg/day. At 12 weeks, add-on CBD reduced median monthly major motor seizures by 50% and total seizures by 44%, with consistent reductions in both seizure types through 96 weeks. At 12 weeks, the proportions of patients with ≥50%, ≥75%, and 100% reductions in major motor seizures were 53%, 23%, and 6%; the proportions with corresponding reductions in total seizures were 46%, 26%, and 5%. Responder rates for both seizure types were consistent through 96 weeks. CBD had an acceptable safety profile; the most common AEs were somnolence (30%) and diarrhea (24%). Conclusions: Results from this interim analysis support add-on CBD as an effective long-term treatment option in LGS or DS.