Aim: The collection of data on longer-term neurodevelopmental outcomes within large neonatal randomized controlled trials by trained assessors can greatly increase costs and present many operational difficulties. The aim of this study was to develop a more practical alternative for identifying major cognitive delay in infants at the age of 24 months, based on parental reports. Method: A sample of 476 infants (206 female, 270 male) previously diagnosed with neonatal sepsis (mean birthweight 1329g [SD 865g], mean gestational age at birth 28.7wks [SD 4.5wks]) from the International Neonatal Immunotherapy Study were assessed using the Parent Report of Children's Abilities--Revised and the Bayley Scales of Infant Development, 2nd edition. Logistic regression was used to model the association between the risk of major cognitive delay (i.e. Bayley Scales of Infant Development Mental Development Index less than 55) and the Parent Report of Children's Abilities--Revised data. Results: The receiver operating characteristic curves for a number of predictive models were constructed--each achieved an area under the curve of at least 90%. The sensitivity, specificity, positive predictive value, and negative predictive value of a number of points on the receiver operating characteristic curves are presented. Interpretation: The Parent Report of Children's Abilities--Revised is a practical tool for identifying major cognitive delay in infants at 24 months.