OBJECTIVE: To describe trends in regionalization of perinatal care and identify factors that predict the extent of regionalization. METHODS: Data were drawn for four states for every year between 1989 and 1998. Panel data models estimated the effect of managed care enrollment on site of delivery for low, very low, and extremely low birth weight neonates. RESULTS: Strong evidence for regionalization over time was observed for North Carolina and Illinois, with little change in site of delivery in Washington. A shift from level III to level II hospitals was observed for low and very low birth weight neonates in California. Although managed care enrollment increased substantially in all four states, managed care had no effect on site of delivery; that is, the effect of managed care was near zero and not statistically significant in any state. CONCLUSION: Evidence supports the delivery of high-risk neonates at tertiary care centers. Despite changes in site of delivery, the percentages of very low birth weight neonates delivered at level III hospitals were substantially lower than the goal of 90% set by Healthy People 2010. Financial pressures introduced by managed care cannot be blamed for the failure to meet this goal. LEVEL OF EVIDENCE: II-2.