In this study, the authors report the results of a three-stage newborn hearing screening (NHS) program for well babies at the Gazi University Hospital (GUH) in Ankara between 2003 and 2013. GUH-NHS was performed by automated transient evoked otoacoustic emission (a-TEOAE) at the first and second steps and by automated brainstem audiometry (a-ABR) at the third step. The data were analysed to assess not only rate of congenital permanent hearing loss (CPHL), but also the effectiveness of the program during the years. A total of 18,470 well babies were tested. The data showed that coverage ratio for the GUH-born babies was increased and more outside-born babies (OBB) were admitted by time (means 84.31 and 11.28 %, respectively). Mean CPHL was found to be 0.26 %. Mean referral rate was decreased to 0.81 % by a-ABR from 2.16 % by a-TEOAE. Mean of missed cases in any stage of GUH-NHS was 4.88 %. It was seen that neither CPHL nor referral rate, but only ratio of missed ones presented increase in parallel to increment in OBB. This paper first presents that clinically acceptable screening procedures developed in GUH by time, and secondly higher rate of CPHL in Turkey than in the Western countries, and benefits of third stage screening by a-ABR because it prevented referral of 251 children (1.29 %) to the clinical tests. We think that this number is reasonably important regarding not only economical point of view, but also waiting lists in the audiology departments in a developing country, in which audiological service is still limited.