Local anesthesia decreases physiologic responses to pain in neonates but has not been used routinely during lumbar punctures in newborns, as it might obscure anatomical landmarks. However, local anesthesia may decrease newborns' struggling during lumbar puncture, thus facilitating the procedure and increasing its success rate. The success rate of lumbar punctures was compared in neonates allocated prospectively to 0.2 to 0.5 mL of 1% lidocaine anesthesia (n = 48) or a control group (n = 52). Newborns were held in a modified lateral recumbent postion (neck not flexed) and their struggling response to the various steps in the lumbar puncture was scored by the holder. The newborns' struggling motion score increased in response to lidocaine injection, but response to the subsequent spinal needle insertion was significantly decreased. Despite this decreased motion, no differences were noted in the number of attempts per lumbar puncture (1.9 ± 0.2 [SEM] in lidocaine and 2.1 ± 0.2 in control groups), rate of lumbar puncture failure (15% in lidocaine and 19% in control groups), or the number of traumatic lumbar punctures (46% in both groups). The success rate of lumbar puncture was not dependent on level of training of physicians performing the procedure. No acute complications, cerebrospinal fluid contamination, or subsequent meningitis was noted in either group. It is concluded that local anesthesia with lidocaine decreases the degree of struggling but does not alter the success rate of lumbar puncture in neonates. The practice of withholding lidocaine anesthesia from neonates undergoing lumbar punctures cannot be justified by arguing that it makes the procedure more difficult to perform.