A combination of the increase in the supply of neonatalogists and competitive pressures lead to the rapid increase in the number of neonatal intensive care units (NICUs) in California in the 1980s. Most of this expansion occurred among intermediate, or level II NICUs and Community NICUs (a California classification for level us that can provide assisted ventilation <6hrs.). Most of these new units and many of the existing level II units were quite small. The net effect of this expansion has been to move a significant percentage of the infants who require NICU care away from tertiary (regional) centers. This has raised concerns about the effects of this expansion on the quality of care. We used the 1990 California birth/infant death cohort file (N=594,104), linked to the hospital discharge abstracts for mothers and infants. Logistic regression was used to examine the effects of patient volume and level of NICU care at the hospital of birth on neonatal mortality, controlling for sex and race of the infant, type of insurance coverage, birth weight, and maternal and neonatal diagnoses associated with neonatal mortality. Compared to level I hospitals, infants born in regional centers had much lower mortality (O.R.-0.75, p=0.003). Both types of level Us had mortality rates that were not significantly different from those at level I units. There was some evidence that community NICUs with more >15 occupied NICU beds had lower mortality, but this result wasn't significant. When costs are used as the dependent variable, there are no significant differences in costs across levels of care, controlling for the other factors in the model. Infants insured by HMOs had mortality rates that were 27% (p=0.015) higher than those of other private insurers. Medicaid mortality was 59% (p=0.0001) higher than those with private insurance. These results indicate that the recent competitive trends in California have had a significant adverse effect on neonatal mortality without realizing any cost savings.