BACKGROUND: Recent clinical findings reported improvement in the treatment outcomes of highly resistant TB (HDR-TB) with the pretomanid (Pa) based regimen. This study aimed to evaluate the cost-effectiveness of the Pa-based regimen for HDR-TB treatment from the perspective of the healthcare sector in the United States.METHODS: A lifelong decision-analytic model was constructed to simulate potential treatment outcomes of 1) the bedaquiline-Pa-linezolid (BPaL) regimen, and 2) the bedaquiline-linezolid (B-L) based regimen in a hypothetical cohort of adult patients with HDR-TB. Primary model outputs were TB-related direct medical costs, qualityadjusted life-years (QALYs) and incremental cost per QALY gained (ICER).RESULTS: In the base-case analysis, the BPaL regimen gained 3.0054 QALYs and saved costs by USD60,433 when compared to the B-L-based regimen. In the probabilistic sensitivity analysis, the BPaL regimen gained higher QALYs at a lower cost in 80.3% of the time, and gained higher QALYs at a higher cost with ICER less than the willingness-to-pay (WTP) threshold (100,000 USD/QALY) in 19.0% of the simulations. The probability of the BPaL regimen being cost-effective was higher than the B-L-based regimen throughout the variation of WTP.CONCLUSION: BPaL therapy is likely the cost-effective option for HDR-TB treatment from the US healthcare sector perspective.