OBJECTIVE The objectives of this study were to 1) quantify the relative importance of established risk factors for congenital hearing loss (HL), 2) identify other risk factors for congenital HL, and 3) create a prognostic system that can predict the chance of an infant having HL. METHODS The authors conducted a retrospective cohort study with validation of a prognostic system. The authors used a medical record review on 1,863 infants admitted to level II and level III nurseries who underwent auditory brainstem reflex (ABR) hearing screening from 1998 to 1999 (derivation cohort). The primary outcome was hearing screening referral (i.e., hearing screening failure) classified as mild to moderate, moderate to severe, and severe to profound loss. To validate the prognostic system, a medical record review of a separate cohort of 437 infants admitted to a level III nursery who underwent ABR hearing screening in 2002 (validation cohort) was performed. RESULTS In the derivation cohort, 1,513 infants (81%) passed the ABR screen at 30 dB, 77 (4%) had a unilateral referral, 243 (13%) had a bilateral referral, and 30 (1.6%) did not have a complete screening. In multivariable analysis, the following risk factors were clinically and statistically significant: craniofacial abnormalities, cytomegalovirus infection, bronchopulmonary dysplasia (BPD), maternal ethanol use, syndromes, hydrocephalus, and hyperbilirubinemia. CONCLUSIONS Our results reaffirm the importance of several established risk factors for congenital HL, but suggest that BPD is an important predictor in high-risk populations. Whether these risk factors are causal or merely associated in congenital HL remains to be determined.