Objective: We studied our collective open-label, compassionate use experience in using cannabidiol (CBD) to treat epilepsy in patients with CDKL5 deficiency disorder and Aicardi, Doose, and Dup15q syndromes. Methods: We included patients aged 1–30 years with severe childhood-onset epilepsy who received CBD for ≥10 weeks as part of multiple investigator-initiated expanded access or state access programs for a compassionate prospective interventional study: CDKL5 deficiency disorder (n = 20), Aicardi syndrome (n = 19), Dup15q syndrome (n = 8), and Doose syndrome (n = 8). These patients were treated at 11 institutions from January 2014 to December 2016. Results: The percent change in median convulsive seizure frequency for all patients taking CBD in the efficacy group decreased from baseline [n = 46] to week 12 (51.4% [n = 35], interquartile range (IQR): 9–85%) and week 48 (59.1% [n = 27], IQR: 14–86%). There was a significant difference between the percent changes in monthly convulsive seizure frequency during baseline and week 12, χ 2 (2) = 22.9, p = 0.00001, with no difference in seizure percent change between weeks 12 and 48. Of the 55 patients in the safety group, 15 (27%) withdrew from extended observation by week 144: 4 due to adverse effects, 9 due to lack of efficacy, 1 withdrew consent, and 1 was lost to follow-up. Significance: This open-label drug trial provides class III evidence for the long-term safety and efficacy of CBD administration in patients with treatment-resistant epilepsy (TRE) associated with CDKL5 deficiency disorder and Aicardi, Dup15q, and Doose syndromes. Adjuvant therapy with CBD showed similar safety and efficacy for these four syndromes as reported in a diverse population of TRE etiologies. This study extended analysis of the prior report from 12 weeks to 48 weeks of efficacy data and suggested that placebo-controlled randomized trials should be conducted to formally assess the safety and efficacy of CBD in these epileptic encephalopathies.