To assess the utility of the CRIB score as a predictor of hospital death and intraventricular hemorrhage (IVH) in very low birth weight (VLBW) and extremely low birth weight (ELBW) neonates. A prospective cohort of VLBW neonates admitted to the neonatal intensive care unit from January 2002 to December 2004 was studied. The data was assessed following the protocol of the SEN 1500 multicenter study. This protocol included assessment of the CRIB score in the first 12 hours of life. Data for the entire group, as well as for two subgroups divided according to birth weight (BW) - VLBW neonates (between 1000 and 1500 g) and ELBW neonates (below 1,000 g) - were evaluated. The area under the receiver operating characteristic curve (Az) was calculated to assess the utility of CRIB score, BW and gestational age (GA). Two multivariate models were used. The cohort consisted of 163 patients. The mean (+/-SD) birthweight was 1.114 (+/-270) g and gestational age (+/-SD) was 29 (+/-3) weeks. The Az for hospital death was 0.757 for the CRIB, 0.758 for BW and 0.703 for GA. The Az for IVH was 0.66 for the CRIB, 0.62 for BW and 0.64 for GA. In the multivariate models for hospital death and IVH, the CRIB was the best predictor. The Az of the CRIB for hospital death was 0.77 for VLBW neonates (p < 0.001) and 0.63 for ELBW neonates (p = 0.82). The predictive utility of the CRIB for hospital death and IVH is similar to that of BW. In the stratification by groups of weight, we found that the CRIB was the best predictor of hospital death in the group weighing > 1,000 g but was no better than chance in the group weighing < 1,000 g.