AIM: The aim of this article was to evaluate the Revised Clinical Risk Index for Babies' (CRIB-II) severity of illness score as a predictor of moderate to severe functional disability (FD) in very premature infants. METHODS: Population study of infants born <29 weeks' gestation cared for in all Neonatal Intensive Care Unit in New South Wales and the Australian Capital Territory between 1998 and 2003. FD at 2-3 years corrected age was defined as developmental delay (quotient < 2 standard deviation), non-ambulatory cerebral palsy (needing aids to walk), blindness (acuity <6/60 in better eye) or deafness (hearing aids). Sensitivity and specificity of CRIB-II scores to predict FD were performed by receiver operating characteristic curve analysis. RESULTS: Of study population of 2210, 480 (21.7%) died before hospital discharge. Among 1328 infants assessed, 217 (16.3%) had FD, 109 (8.2%) developmental delay, 75 (5.6%) cerebral palsy and 54 (4.1%) blindness or deafness. CRIB-II performed significantly better than gestation or birthweight (BW) alone in predicting mortality (area under the curve (AUC) ± standard error 0.83 ± 0.01, vs. 0.78 ± 0.01 and 0.76 ± 0.01, respectively). CRIB-II scores were significantly higher in FD than non-FD children (11.9 ± 2.9 vs. 10.1 ± 2.6), but the AUC for CRIB-II (0.68 ± 0.02) did not significantly differ from that of gestation (0.65 ± 0.02) and BW (0.65 ± 0.02). CONCLUSION: CRIB-II improved prediction of mortality but did not perform better than gestational age or BW in predicting FD. We would caution clinicians against using the infant's condition at admission to predict long-term outcome.