Epidural anaesthesia (EA) is the most important analgesia technique in obstetrics for delivery. In pregnancy, hormonal adjustments lead to an alteration of tissue consistency, which often causes an early, untimely loss of resistance. Apart from mere inspection and palpation, no useful diagnostic method prior to EA performance has been established yet. In this prospective study, we examined 100 pregnant women, who had been admitted for childbirth and were undergoing epidural block (level L3-L4) for delivery. Sonotopography of the lumbar epidural structures was performed directly before epidural puncture and childbirth. We evaluated the visibility of all anatomical structures and compared all distances measured by ultrasonography and during puncture. The correlation between distances measured by ultrasound and by puncture needle was high (r2 = 0.79). No obvious dependency was found between ultrasonic and puncture angle (r2 = 0.19). The temporal distance from ultrasonic examination and puncture causes unavoidable differences: each deviation between ultrasound and puncture conditions causes a modification of the puncture depth. The patient acceptance of the procedure was very good. Ultrasonography offers the possibility to determine site and direction of epidural puncture and distance of the epidural space to the skin even before the puncture attempt. The ultrasound controlled EA for delivery can easily be inserted into the clinical routine. Ultrasonography can fill an important diagnostic gap in regional anaesthesia.