In a previous study of very low birth weight neonates, < or = 1500 g, admitted to the Vanderbilt University Neonatal Intensive Care Unit (NICU) from 1976-1990, improvements in survival were accompanied by a corresponding increase in the incidence of bronchopulmonary dysplasia (BPD). Since then, certain neonatal and perinatal interventions have been introduced and may influence neonatal outcomes. In this study, we have continued the analysis of the incidence of 3 outcomes: 1) Neonatal death (NEOD), 2) BPD, and 3) NEOD or BPD (NEOD/BPD) for an additional 7 years, 1991-1997. A retrospective study was performed of 3,837 patients with birth weight < or = 1500 g and admitted to the Vanderbilt NICU within 24 hours of birth from 1976 through 1997. The outcomes NEOD, BPD, or NEOD/BPD were modeled by using multiple logistic regression with the following risk factors included as covariates: birth weight, gestational age, Apgar scores at 1 and 5 minutes, gender, race, birth location, diagnosis of hyaline membrane disease, maternal age, maternal diabetes, delivery method, multiple births, duration of ruptured membranes, and biologically relevant interactions among these covariates. To assess time trends in the risk factors and outcomes, patients were divided into time periods (1 = 1976-80, 2 = 1981-85, 3 = 1986-90, 4 = 1991-95, and 5 = 1996-97). For each outcome, only covariates or interactions among covariates found to be significant were retained in the final model. Adjusted odds ratios and 95% confidence intervals were calculated to measure the risk associated with a given time period in comparison to the preceding period. There was a progressive decline in NEOD across all time periods. The previously described increase in BPD from period 1 through period 3 is followed by a decrease in periods 4 and 5. The risk of NEOD/BPD remained fairly constant from period 1 to period 3, but then showed a significant decrease over the two most recent periods. Prior to 1991, the cost of improved survival among very low birth weight infants in this large NICU was an increased incidence of BPD. Since 1991, the risk of BPD has been decreasing even though survival continues to improve. If these findings are also representative of other NICUs, they signify an important reduction in the impact of BPD as one of the costly sequelae of prematurity.