This study was designed to investigate the effect of steroid administration in ill premature neonates. Twenty high-risk very-low-birth-weight (VLBW) infants [birth weight (BW) < or = 1,300 g] with a mean BW 948 +/- 220 g, gestational age (GA) 27 +/- 1.7 weeks underwent 1-hour ACTH (Cortrosyn) stimulation tests and determination of 17-hydroxyprogesterone (17OHP)/dehydroepiandrosterone sulfate (DHEAS) at 23.6 +/- 15.9 days poststeroid treatment for bronchopulmonary dysplasia (BPD)/airway obstruction. Metyrapone tests were also obtained in 18 infants. Baseline (nonsteroid-exposed) values for pre-/post-ACTH cortisol, 17OHP, DHEAS, and pre-/post-metyrapone compound S values were obtained in 5 infants. Eight of 18 (44%) infants had evidence of secondary (hypothalamic-pituitary) adrenal suppression based on abnormal metyrapone tests. No difference was found in BW, GA, time on O2 or AV, steroid dose/kg, or neonatal/postneonatal mortality between the suppressed and nonsuppressed groups. Two of 4 infants with borderline ACTH tests had subnormal compound S levels postmetyrapone. No relationship was found between steroid dose/kg and cortisol response post-ACTH. Additionally, corrected GA was not related to change in cortisol, 17OHP, and DHEAS pre-/post-ACTH. Two infants exhibited recovery of adrenal suppression documented by repeated metyrapone testing at 63 and 186 days poststeroid treatment. In conclusion, this study documents the apparent high incidence of secondary adrenal suppression in VLBW infants treated with dexamethasone. Clinical significance of these findings deserves further investigation.