OBJECTIVES: To examine the cost-effectiveness of adding nucleic acid testing (NAT) to enzyme immunoassay (EIA) screening for hepatitis B (HBV), hepatitis C (HCV), and human immunodeficiency virus (HIV) in Taiwan based on the governmental perspective. METHODS: A Markov model depicting the natural history of HBV, HCV, and HIV infections and subsequent illnesses for blood recipients in Taiwan was developed to evaluate the cost-effectiveness of adding five different NAT screening strategies employing different test systems and pool sizes to EIA screening, including cobas TaqScreen multiplex in pools of 6 (EIA+Cobas6), Procelix Ultrio in individual donations (EIA+Ultrio1), or pools of 4 (EIA+Ultrio4), and Procelix Ultrio Plus in individual donations (EIA+UPlus1), or pools of 8 (EIA+UPlus8). The population size and characteristics of the recipients were derived from the National Health Insurance (NHI) data. Transition probabilities and utilities were abstracted from published literature. Efficacy and screening costs were obtained from a nationwide NAT screening program launched in Taiwan. Medical costs incurred by infections and subsequent illnesses were obtained from the NHI and published literature based on Taiwanese data. Both costs and health outcomes were discounted at 3%. All costs were presented in 2012 USD. Probabilistic sensitivity analyses were performed. RESULTS: A 270,535-recipient cohort based on the annual number of recipients in Taiwan, was simulated. The model predicted that for various NAT screening strategies 120-400 infections were averted with additional 6.8-13.6 million NAT screening costs, corresponding to 34,200-72,400 USD per infection averted, and the lifetime incremental costs per quality-adjusted life-year (QALY) gained were 48,600-72,400 USD/QALY. Wide variation, however, was seen and the probability that NAT being cost-effective was 18-30%. CONCLUSIONS: EIA+Cobas6 and EIA+UPlus8 are the two strategies of better cost-effectiveness. At the threshold of three-fold GDP per capita (55,800 USD/QALY), EIA+Cobas6 and EIA+UPlus8 may be considered as cost-effective in Taiwan but with high uncertainty.