The records for all children from a five-year birth cohort in the geographically defined area of Lothian who were referred by their health visitor to a centralized second tier audiology service after they failed their 7-9 month infant distraction hearing screen were reviewed. The sensitivities, specificities, positive predictive values and yield of this screen for the detection of significant sensorineural hearing loss > or =40 dBA requiring aiding and for a conductive loss persisting beyond one month over 30 dBA were determined. In addition, the records were studied for the same five-year birth cohort for children presenting to the centralized multidisciplinary third tier clinic with a significant sensorineural hearing loss requiring aiding, and the reasons for late diagnosis were determined. If high risk neonatal screening had been available for this population then potentially 67.5% (27 out of 40) of cases could have been identified by an appropriate age. Elimination of health visitor distraction screen false negatives for cases with loses over 60 dBA would potentially increase the case identification by the appropriate age to 45% (1 8 out of 40). Thus, there is greater scope for improving our present results with the introduction of high risk neonatal screening than by improvements in the health visitor screen.