• Regulatory approval of new tuberculosis (TB) drugs can be based on data from trial(s) using a surrogate endpoint of treatment efficacy under an accelerated or conditional procedure. In such circumstances, policy makers and TB programs can be hampered in their ability to make recommendations on the optimal use of the drug(s), and consequently , the uptake by national or international public health institutions of such recommendations can be limited. • Based on the essential need to produce high-quality evidence for policy decisions, this paper reflects on specific methodological issues in clinical trial design that need to be addressed to improve compliance with clinical, regulatory, and public health requirements. • Established mechanisms for communication between drug developers and regulators already exist; however, equal engagement with policy makers is also essential for the optimal selection of trial designs, endpoints, and markers of treatment outcome and for giving consideration to public health and program aspects. • The next generation of TB trials should better reconcile the research agenda with the need for global policies on access to TB medicines. Policy decision-makers should establish formal mechanisms for iterative feedback on regimen-development pathways. In this paper, we provide examples of how the need for interactions between regulators, tri-alists, and policy decision-makers can be addressed.