Objective: To examine temporal trends in the rates of severe bronchopulmonary dysplasia (BPD) between 1994 and 2002. Study design: In a retrospective cohort study, all infants with a gestational age (GA) < 33 weeks in a large managed car e organization were identified. Annual rates of BPD (defined as an oxygen requir ement at 36 weeks corrected GA), severe BPD (defined as respiratory support at 3 6 weeks corrected GA), and death before 36 weeks corrected GA were examined. Res ults: Of the 5115 infants in the study cohort, 603 (12% ) had BPD, including 24 6 (4.9% ) who had severe BPD. There were 481 (9.5% ) deaths before 36 weeks co rrected GA. Although the decline in BPD in this period was not significant, the rates of severe BPD declined from 9.7% in 1994 to 3.7% in 2002. Controlling for gestational age, the odds ratio (95% CI) for annual rate of decline in severe BPD was 0.890 (0.841- 0.941). Controlling for gestational ag e, deaths before 36 weeks corrected GA also declined, with the odds ratio (CI) f or the annual decline being 0.944 (0.896- 0.996). Conclusions: In this study po pulation, the odds of having of BPD remained constant after controlling for GA. However, the odds of having severe BPD declined on average 11% per year betwee n 1994 and 2002. Objective: To examine temporal trends in the rates of severe bronchopulmonary dysplasia (BPD) between 1994 and 2002. Study design: In a retrospective cohort study, all infants with a gestational age (GA) <33 weeks in a large managed car organization identified. Annual rates of BPD (defined as an oxygen requir ement at 36 weeks corrected GA), severe BPD (defined as respiratory support at 3 6 weeks corrected GA), and died before 36 weeks corrected GA were examined. Res ults: Of the 5115 infants in the study cohort, 603 (12%) had BPD, including 24 6 (4.9%) who had severe BPD. There were 481 (9.5%) deaths before 36 weeks co rrected GA. was not significant, the rates of severe BPD declined from 9.7% in 1994 to 3.7% in 2002. Controlling for gestational age, the odds ratio (95% CI) for annual rate of decline in severe BPD was 0.890 (0.841-0.941). Controlling for gestational ag e, deaths before 36 weeks corrected GA also declined, with th The odds ratio (CI) f or the annual decline being 0.944 (0.896- 0.996). Conclusions: In this study po pulation, the odds of having of BPD remained constant after controlling for GA. However, the odds of having severe BPD declined on average 11% per year betwee n 1994 and 2002.