Thirty patients presenting with post-dural puncture headache (PDPH) were prospectively studied to determine the influence of the duration of the decubitus position after epidural blood patch on the efficacy of treatment. All patients received 12 ml of autologous blood. They were randomly distributed into three groups of ten patients. Patients in Group 1 were maintained in a decubitus position for 30 min after the epidural injection of autologous blood in the epidural space. Patients in Group 2 were maintained for 60 min in decubitus and patients in Group 3 for 120 min. Post-dural puncture headache was evaluated using a visual analogue scale before the epidural blood patch, at the time of initially adopting a standing position after the blood patch, and 24 hr later. The severity of PDPH in the three groups was reduced at the time of initially adopting a standing position and after 24 hr, in comparison with preblood patch VAS (P < 0.001). Patients in Group 3 presented less severe PDPH than patients in Group 1 at the time of initially standing up and 24 hr later (P < 0.05). We conclude that epidural blood patch was effective in treating PDPH but that the maintenance of a decubitus position for at least one hour and preferably for two hours after the blood patch was more effective than maintenance for 30 min.