While postnatal dexamethasone (dex) therapy ameliorates bronchopulmonary dysplasia (BPD) in preterm infants, it also increases rates of neuromotor and cognitive abnormalities. In a study by Murphy et al, dex therapy was associated with reduced cerebral cortical gray matter (GM) volumes as compared to untreated infants on volumetric MRI.1However, in that study, the impact of BPD was not considered as a confounder and dex therapy was given for a prolonged period (mean duration, 28 days). We evaluated brain component volumes in a cohort of extremely low birth weight infants (ELBW, BW ≤ 1000 g) discharged from our neonatal intensive care unit during a 7-month period, from June 2003 through December 2003. We hypothesized that dex therapy results in reduced combined cortical gray and white matter (WM) volume (cortical volume) with concomitant increased total cerebrospinal fluid (CSF) volume, at a given total brain volume. Of the 38 ELBW infants who survived to discharge, 37 had an anatomical brain MRI at term corrected age. Twenty of the 37 infants (54%) received dex for ≤ 7 days, beginning after two weeks of age for evolving BPD. MRIs were segmented using semiautomated scoring algorithms to generate 3-dimensional brain volumes. Analyses are complete for 19 infants (n = 11 dex-treated; n = 8 untreated). On univariate analysis, dex-treated infants demonstrated a nonsignificant 4.4% reduction in cortical volume (p = .23) and a 13.7% increase in total CSF volume as compared to untreated infants (p = .26). Cortical GM volumes were similar in the two groups (p = .61). Cortical WM was significantly reduced by 23.4% in the dex group as compared to the untreated group (p = .03). This difference in cortical WM persisted after adjustment for multiple potential confounders, including BPD and birth weight. In ELBW infants, postnatal dex therapy after two weeks of life, for a duration of 7 days or less, may adversely affect developing white matter growth, possibly contributing to increased rates of cerebral palsy and other neuromotor abnormalities. Randomized trials of other anti-inflammatory agents with assessment of brain volumes and neurodevelopmental outcomes are urgently needed in this population of vulnerable infants.