Objective: Hong Kong is a high-income city of China with an intermediate tuberculosis (TB) burden, and 1% of TB cases are multidrug-resistant (MDR-TB). The aim of this study was to examine the potential costeffectiveness of adding bedaquiline or delamanid to the background regimen (BR) for the treatment of MDR-TB in Hong Kong. Methods: A decision-analytic model was designed to simulate outcomes over a 10-year time horizon for MDR-TB patients treated with bedaquiline plus BR (B-BR), delamanid plus BR (D-BR), or BR alone. Outcome measures included direct medical costs and quality-adjusted life-years (QALYs) gained. Results: In the base-case analysis, BR was the least costly regimen (USD 47 396) with the lowest QALYs gained (6.347). Compared to BR, B-BR gained an additional 0.731 QALYs with incremental cost of USD 9. The incremental cost-effectiveness ratio (ICER) of B-BR was USD 12/QALY. D-BR was more costly than BR by USD 20 164 and gained an additional 0.012 QALYs. The ICER of D-BR was USD 1 680 333/QALY. In the probabilistic sensitivity analysis with 10 000 Monte Carlo simulations, B-BR and D-BR were cost-effective 99.98% and 5.13% of the time, respectively, using 1Â gross domestic product per capita (USD 46 182) as the willingness-to-pay threshold. Conclusions: Bedaquiline is more likely than delamanid to be cost-effective when added to BR for the treatment of MDR-TB in Hong Kong.