OBJECTIVES: According to WHO the use of antenatal corticosteroids (CEA) in pregnant women at risk of preterm birth <34 weeks can prevent thousands of preterm neonates (PN) deaths. The impact of the use of CEA in hospital costs in developing countries is not known. Our objective was to compare morbidity and hospital costs of PN whose mothers received or not CEA.  METHODS: Analysis of PN medical records with gestational age 26-32 weeks born from Jan/2006-Dez/2009 in a tertiary, public and university hospital. We excluded infants with malformations. Maternal characteristics, hospital neonatal morbidity, use and doses of CEA and all used resources (tests, medications and procedures) were collected. Costs were estimated in Brazilian Reais, from the hospital perspective. RESULTS: Of 211 PN, 170 received at least one dose of CEA to 6 hours before delivery (G1) and 41 did not (G2). The groups had similar characteristics but G1 had more male infants (p <0.05) and cesarean sections (p <0,00). Morbidity: G2 needed more advanced resuscitation (16.5% vs 34%, p = 0.01), experienced more intraventricular hemorrhage III / IV (7.6% vs. 22%, p <0.00) and retinopathy of prematurity (12.4% vs. 24.4%, p = 0.05). Resource use: G1 consumed less mechanical ventilation days (5.3 vs 10.6, p = 0.04) and oxygen days (10.7 vs 17, p = 0.02); the number of NICU and Intermediate Care Nursery days were respectively (19.6 vs 27.5, p = 0.07) and (24 vs 29.5 days, p = 0.14); there was no difference concerning use of CPAP (p = 0.07) and surfactant (p = 0.06). The average cost of hospitalization per patient was BRL 18,409 in G1 and BRL 24,090 in G2 (p = 0.03).  CONCLUSIONS: The CEA is a simple measure, which helps to reduce PN morbidity and utilization of health care resources, reducing hospital costs.