Aim Evaluation of automated auditory brainstem response (AABR) hearing screening in Dutch neonatal intensive care units (NICU). Design As part of a nationwide hearing screening programme two-stage AABR hearing screening has gradually been introduced in all Dutch NICU. From 1998 to 2007 graduates with more than one risk factors according to the 2000 Joint Committee on Infant Hearing statement were included. Central registration of results facilitates screening, tracking and follow-up after abnormal screening results. Quality assurance of the programme was established. Endpoints are prevalence of uni/bilateral hearing loss and referral rate after the 1st and 2nd test. Quality indicators are: participation rate 1st test (goal >98%), 2nd test (goal >95%), diagnostic examination (goal >95%) and percentage of neonates with an age corrected for gestational age at the 1st test <30 days (goal >90%), 2nd test <5 weeks (>90%), diagnosis <13 weeks (goal >90%). Results 29 110 neonates have been screened. 9.2% did not pass the 1st test. After the 2nd test, 2.7% (n = 778) were referred for diagnostic examination. Bilateral hearing loss was diagnosed in 61% of the referrals (prevalence 1.6%) and unilateral hearing loss in 17% (prevalence 0.5%). The participation rates over years varied between 98.0 and 99.6% (1st test), 88.5 and 94.4% (2nd test), 88.6 and 6.5% (diagnostic examination). 96% of the neonates had their 1st test at <30 days and 77% of the referred neonates had their 2nd test at <5 weeks. 67% of the referred neonates were diagnosed at <13 weeks. Conclusion The AABR neonatal hearing screening programme in Dutch NICU is highly effective in terms of participation and referral rate. The positive predictive value of the two-stage programme is 79%.