Obstetric patients have a higher incidence of postdural puncture headache (PDPH) than the corresponding incidence in the general population. Current evidence suggests that pencil-point spinal needles are significantly superior compared to cutting spinal needles regarding the frequency of PDPH [1,2]. Moreover, the impact of needle gauge on PDPH has been shown to be less important in pencil-point as compared to cutting needles [3]. Atraucan is a double-beveled needle containing cutting and blunt components of the bevel. It was categorized as a non-traumatic needle in a recent meta-analysis [1] but its place regarding PDPH is not fully elucidated. Introducing cutting needles with the bevel parallel to the longitudinal dural fibers is found to be a factor associated with reduction of the incidence of PDPH [4].The notion that the paramedian approach decreases the incidence of PDPH is supported by studies in models of the dura mater in vitro showing that the oblique angle of dural puncture causes less leakage across the dura [5,6]. Clinical studies regarding the influence of the paramedian approach on PDPH have shown contradictory results.The association between the patient's position during lumbar puncture and PDPH was the object of Abstract Background: Concern has been raised that Sprotte needles predispose to spinal anes thesia failure. Nevertheless, these needles are associated with a low incidence of post dural puncture headache. The impact of the paramedian approach to postdural punc ture headache remains controversial. The objective of this prospective randomized study was to compare Sprotte, Quincke and Atraucan needles as well as the midline and the paramedian approach in terms of postdural puncture headache and spinal anesthesia failure in patients undergoing Caesarean section.Methods: 655 patients were randomized to 5 groups. A midline approach was used in four groups. The spinal needles were the 25G Sprotte, 27G Sprotte, 26G Atraucan and 25G Quincke. In the fifth group a 25G Quincke needle was used by the paramedian approach.Results: The incidence of postdural puncture headache was 0% in both 25G and 27G Sprotte groups, 2.5% in the 26G Atraucan group, and 7.2% and 2.7% in the 25G Quincke midline and paramedian approach respectively. A significant difference in terms of post dural puncture headache was found between 25G Sprotte and 25G Quincke needles (P = 0.004), while the failure rate was similar between these two needles. A significant difference in spinal anesthesia failure rate was observed between midline and parame dian approaches (P = 0.041).Conclusions: Sprotte but not Atraucan needle design correlates with lower incidence of postdural puncture headache compared to Quincke design. Sprotte needles are not associated with a higher spinal anesthesia failure compared to Quincke needles. The incidence of postdural puncture headache by the paramedian approach is not sig nificantly reduced whereas the spinal anesthesia failure rate is increased in comparison to the midline approach.