Post dural puncture headache (PDPH) represents a complication of anesthesia (with an increased incidence in obstetric patients) or as the consequence of a diagnostic lumbar puncture. The aim of the present study was to evaluate the efficacy of the epidural blood-patch (EBP) versus the conventional medical treatment of post-anesthetic headaches also including the PDPH following a diagnostic puncture, a category of patients rarely referred to the anesthesia consultation in our hospital because it was believed that they might have equal benefit from conventional measures due to the smaller size of needles used. We studied in a prospective, randomized, double-blinded manner 32 obstetric and non-obstetric patients with PDPH having the onset of the symptoms 24 hours before the inclusion in the study. The patients were randomly divided in two groups: group A (16 patients) receiving conventional treatment (oral and intravenous fluid replacement, non-steroidal anti-inflammatory drugs--NSAIDs--, caffeine) and group B (16 patients) in whom an epidural blood-patch was performed. The intensity of the headache was evaluated using a visual analogue scale (VAS) from 0-10, before, 2 hours and 24 hours after the EBP. There were no statistical differences concerning the demographic data and the cause of PDPH between the groups (p > 0.05). The intensity of PDPH was similar before performing the EBP (p > 0.05), with a value on VAS of 8.2 +/- 1,4. in group A and 8,0 +/- 1.6 in group B. Two hours after the treatment, the intensity of headache on VAS diminished extremely significant (p < 0.0001): in group B the value was 1.0 +/- 0,18 versus 8.2 +/- 1.4 in group A. The difference recorded after 24 hours remained statistically significant (p < 0.0001): the VAS scores were 0.7 +/- 0,16 and 7.8 +/- 1.2 respectively. The epidural blood patch represents the first choice treatment of PDPH no matter the etiology, being significantly superior to the conventional treatment which did not affect pain scores. In severe PDPH there is no reason to delay the EBP more than 24 hours after the diagnosis as all except two patients of the conventional treatment group required blood patching following the study period.