We hypothesize that the respiratory and functional states of the very low birth weight infant with bronchopulmonary dysplasia can be improved in the neonatal intensive care unit by prevention of inappropriate sensory input. To test this hypothesis, we developed for preterm newborns a behavior observation method that catalogues specific reaction patterns according to putative stress and relaxation behaviors. We then collected behavioral information and heart rate, respiratory rate, and transcutaneous PO2 readings before, during, and after routine care-giving interventions. Eight control and eight experimental infants were selected for study based on the following criteria: birth weight <1,250 g, gestational age <28 weeks, on the respirator >24 hours in first 48 hours of life at ≥0.6O FiO2 for more than two hours during first 48 hours of life. Additionally, the two groups were comparable on other medical and demographic variables, including severity of respiratory status for the first ten days and incidence of intraventricular hemorrhage, patent ductus arteriosus, and socioeconomic status. Systematic observations were conducted on days 10, 20, and 30 after birth and at 36 and 40 weeks postconception. For the intervention infants, our observations were discussed with the infants' primary nurses, and individualized modifications for each infant's care plan were implemented based on these observations. Experimental infants showed significantly briefer stays on the respirator (P < .01) and in increased FiO2 (P < .05). Their feeding behavior was normalized significantly earlier (P < .01). Experimental infants also showed significantly better behavioral regulation scores at 1 month after their mothers' estimated dates of confinement (post-EDC), as measured with the Assessment of Preterm Infants' Behavior, significantly better Mental and Psychomotor Developmental Indices at 3, 6, and 9 months post-EDC, as measured with the Bayley Scales of Infant Development, and significantly better behavioral regulation scores at 9 months post-EDC, as measured in a videotaped play observation. Measurements of weight, height, and head circumference at 3, 6, and 9 months post-EDC showed no differences. All assessments were performed by one of two trained testers not familiar with the goals of the study or the group status of the infant. These results support the hypothesis that very low birth weight preterm babies profit significantly both medically and developmentally from individualized behavioral care in the neonatal intensive care unit.