Lumbar puncture is an essential part of a complete sepsis evaluation in the newborn period. It is a procedure that can be difficult to perform and may yield data that are difficult to interpret. This has raised questions about the indications for lumbar puncture in newborns. Objective: To investigate the success rates for lumbar punctures completed in a Level III NICU as the first step in improving the outcomes for the procedure. Methods: Retrospective chart review to include all lumbar punctures that were completed and documented during 1995. A successful procedure was defined as one in which CSF was obtained. Results: Overall, 355 lumbar punctures were completed and documented in 1995. 155/355 were successful (44%). Success rates as determined by the individual performing the procedure were as follows: Attending/fellow neonatologist- 44/84 (52%), neonatal nurse practitioner-38/77 (49%), resident-73/194 (38%). Success rates based on gestational age were: preterm infants(<36 wks)- 95/183 (52%), term infants (>36 wks)- 60/172 (35%). Neonatologists were more successful than residents (Chi-square p=.02), but there was no difference between neonatologists and NNPs or NNPs and residents (Chi-square p=.70 and.07, respectively). However, a low success rate (52% at best) was independent of the neonatal provider for the procedure. There was a better outcome for preterm v. term patients (Chi-square p=.001). Conclusions: This data suggest that factors beyond experience and training of the provider contribute to the outcome in lumbar puncture. A review of procedure technique, including patient positioning and pain management, is indicated.