Objective: To evaluate the usefulness of two-site blood cultures for the documentation of bacterial clearance during initial treatment of culture proven neonatal sepsis. Design/methods: Clinical data were prospectively collected for 216 neonates who had blood cultures drawn for evaluation of possible sepsis. A positive diagnostic blood culture was followed by repeat blood cultures from two different peripheral sites after initiation of antimicrobial therapy. Results: Follow-up two-site cultures were drawn on 25 occasions at a median interval of 3 days after start of antimicrobial therapy for culture-proven sepsis in 15 neonates. On 12 occasions, follow-up blood cultures were both negative. On eight occasions, follow-up blood cultures grew the same organisms from both sites. However, follow-up two-site cultures on five occasions in four of the 15 infants grew organisms from one site but not the second site. Conclusions: Two-site blood cultures seem necessary to document bacterial clearance in response to antimicrobial therapy for culture-proven sepsis, and may help determine the appropriate duration of antimicrobial therapy by reducing false-negative results.