Background: Neonatal sepsis is a cause of mortality and long-term morbidity worldwide. Objectives: To describe longitudinal trends in the cumulative incidence of early-and late-onset sepsis (EOS and LOS), mortality, and causative organisms in a Thai Hospital before and after construction of a new neonatal intensive care unit (NICU). Methods: Review of NICU admissions with blood cultures positive for bacteria or fungi for the periods 1995 to 2002 (preconstruction) and 2004 to 2010 (postconstruction). Sepsis was categorized into EOS (within first 3 days of life) and LOS (after first 3 days of life). Results: Of 5,570 admissions, 241 (4.3%) neonates with 276 episodes of sepsis were identified. There was no difference in the rate of sepsis overall (P = 0.90), LOS (P = 0.30), or sepsis-related mortality (P = 0.61) over the two periods, but the rate of EOS increased significantly from 0.34% to 0.81% (P = 0.04). Rates of Klebsiella species and Escherichia coli sepsis increased from 13.6% to 25.6% (P = 0.01) and from 5.3% to 12.2% (P = 0.04), respectively, while rates of Staphylococcus aureus sepsis decreased from 12.9% to 4.3% (P < 0.007). Sepsisrelated mortality was 1.8%. Conclusions: Although direct causality cannot be proven, the rate of EOS and the pattern of causative organisms changed following construction of the new NICU. Building a new unit does not necessarily result in a reduction in the rate of sepsis. This data may provide a baseline for implementing evidence-based infection control strategies to prevent/reduce sepsis and improve neonatal care.