Cervicogenic headache (CEH) is caused by a structural abnormality in the cervical spine. Available treatments for CEH include medical therapy, local botulinum toxin injection, cervical epidural corticosteroid injection, and surgery. The objective of this study was to investigate the safety and efficacy of a continuous epidural block of the cervical vertebra. Medical records were retrospectively analyzed for 37 patients diagnosed with CEH treated by a continuous epidural block of the cervical vertebra with lidocaine, dexamethasone, and saline (5 ml/min) for 3 - 4 weeks and triamcinolone acetonide 5 mg once weekly for 3 - 4 weeks. Pain was measured via the visual analogue scale (VAS) in combination with quality of life assessment. Outcome measures were patient-reported days with mild or moderate pain, occurrence of severe pain, and the daily oral dosages of non-steroidal anti-inflammatory drug use (NSAID). In the 3 months immediately preceding placement of the epidural catheter, the mean number of days with mild or moderate pain was 22.0 +/- 4.3. The mean occurrence of severe pain was (3.20 +/- 0.75) times and the mean oral dosage of NSAID was (1267 +/- 325) mg. During the first 6 months after epidural administration of lidocaine and corticosteroids, the mean number of days with mild or moderate pain, the mean occurrence of severe pain, and the mean daily oral dosages of NSAIDs were significantly decreased compared to 3-month period immediately preceding treatment (P < 0.01). By 12 months post-treatment, no significant difference in these three outcome measures was noted. Continuous epidural block of the cervical vertebra for patients with CEH is effective for at least six months. Further research is needed to elucidate mechanisms of action and to prolong this effect.