Early identification of hearing impairment may improve language outcomes and subsequent school and occupational performance of the deaf. Universal newborn hearing screening (UNHS), currently mandated by 32 states, can reduce the median age of identification of hearing impairment from 12 to 18 months to 6 months or less. However, because false-negative tests must be minimized, the prevalence of congenital deafness is low, and screening tests are imperfect, UNHS results in many false-positive results and has a low positive predictive value (PPV). The objective of this study was to evaluate UNHS and selective screening in terms of both short- and long-term benefits, harms, and financial costs and to identify steps in the screening process that could be improved to increase cost-effectiveness. The cost-effectiveness analysis, conducted from the societal perspective, compared the projected outcomes of 1) no newborn hearing screening, 2) selective newborn hearing screening, and 3) UNHS for a hypothetical state birth cohort of 80 000 infants. Probability and cost estimates for the decision model were obtained from published studies, expert opinion, and national and state sources. The main outcomes were incremental cost per infant whose deafness was diagnosed by 6 months, which included only the cost of screening and diagnostic evaluation; and incremental cost per deaf child with normal language, which also included the costs of medical care, education and assistive devices, and lost productivity over the lifetime of the deaf individual. Selective screening identified 62 of the 128 deaf infants in the birth cohort, referred 0.18% of all infants for diagnostic evaluation, and had a PPV of 43%. UNHS identified 116 of the 128 deaf infants, referred 1.6% of all infants, and had a PPV of 8.8%. Our model simulated real-world conditions in which some infants whose deafness is identified at screening do not receive a definitive diagnosis of being deaf before 6 months; and a portion of deaf and hard-of-hearing infants who 1) have false-negative screening test results, 2) are not screened, or 3) fail the hearing screen but are not immediately followed up with diagnostic evaluation nonetheless receive a diagnosis by 6 months of age. In the absence of newborn hearing screening, approximately 30 deaf infants were identified by 6 months of age by passive detection alone at a cost of $69 000. The selective screening protocol, when compared with no newborn hearing screening, resulted in an additional 36 infants whose deafness was diagnosed by 6 months at an additional cost of approximately $600 000, yielding an incremental cost-effectiveness of approximately $16 000 per additional infant whose deafness was diagnosed by 6 months. Compared with selective screening, the UNHS protocol resulted in 33 additional infants whose deafness was diagnosed by 6 months of age at an additional cost of approximately $1.5 million, yielding an incremental cost-effectiveness of approximately $44 000 per additional infant whose deafness was diagnosed by 6 months of age. Increasing the rate of follow-up to diagnostic evaluation from the base-case estimate of 77% to 100% decreased the incremental cost of UNHS to $38 000 per additional infant whose deafness was diagnosed by 6 months. Under the base-case assumptions about lifetime savings that result from normal language with early intervention, UNHS resulted in normal language achievement for more deaf children and was cost saving in the long term compared with both selective screening and no screening. The short-term cost-effectiveness of UNHS is comparable to the cost per case diagnosed of other newborn screening programs and could be improved by increasing the rate of follow-up to diagnostic evaluation after positive screening test results. If early identification results in improved language abilities, lower educational and vocational costs, and increased lifetime productivity, then UNHS has the potential for long-term cost savings compared with selective hearing screening and no screening. To understand the actual long-term economic effects of UNHS, better evidence is needed regarding the impact of early intervention on language outcomes and subsequent changes in educational costs and lifetime productivity.