Objective: We sought to evaluate the agreement between the opinions of specialists in fetal medicine with ex vivo observations on the potential influence of the angle of needle insertion on the fluid leak through the created defect in human fetal membranes. Study Design: Membranes from placentas of women who were delivered by elective cesarean were harvested, cut in pieces, and secured to the bottom of plastic tubes filled with Hartmann solution. They were punctured with 18-, 20-, or 22-gauge needles, with an angle of insertion of 90° (group 1) or 45° (group 2), and the flow rate (in milliliters per minute) through the created defect at a constant pressure of 150 mm H₂O was measured. Fifty physicians performing amniocentesis at fetal medicine reference centers were interviewed about their impression and clinical attitude with respect to the angle of needle insertion at the time of amniocentesis. Results: In the ex vivo study, puncture with a 45° angle was associated with a significantly lower flow of fluid through the membrane defect for all needle sizes tested. Regarding survey answers, 82% of physicians try to perform amniocentesis with a given angle (ie, 90° in the vast majority of cases). Among the reasons for doing so, minimizing membrane damage was mentioned in more than half of cases. Conclusion: These results provide evidence that the angle of needle insertion influences the type of defect and therefore the flow rate through human membranes. Membrane damage is a concern of a proportion of specialists while performing amniocentesis, but ex vivo observations do not agree with clinical assumptions on the potential influence of the angle of insertion. (Am J Obstet Gynecol 2000;182:646-9.)