Drug development in TB requires new integrated methods to transition the novel combination regimens needed to shorten first-line therapy and combat multi-drug resistance. Although new agents are emerging, the path to registration of such regimens remains uncertain while capacity for pivotal trials is limited. Selection and optimization of drug combinations for development depends on preclinical systems which do not capture in vivo pharmacodynamics of heterogeneous states of M.tuberculosis.These systems are diverse and their predictive power for clinical trial outcomes is uncertain, escalating risks during development. To overcome these bottlenecks, PreDiCT-TB will take a comprehensive model based approach, synthesizing and integrating preclinical and clinical information. We will harness innovative technologies to develop an integrated set of predictive pre-clinical tools to facilitate selection of the best drug combinations and dosages for entry into optimized clinical studies. We will develop diverse existing and innovative experimental models of pharmacodynamics from extracellular MTB organisms, through to acute and chronic animal models in different species. We will enhance these systems using innovative single cell and functional imaging, novel mycobacteriological approaches and identification of biomarkers of lethality and sterilization. We will represent these critical data using multi-scale mathematical approaches incorporating PK-PD and disease models, supported by a translational data integration platform.The aim is an optimized decision pathway for combination regimens to progress to innovative early clinical trials. We will develop this modelling framework using a training set of reference compounds and subsequently update it with emerging data from new compounds. We believe that this comprehensive approach is the only way to overcome existing gaps in translation and create an effective and rapid drug development strategy for European researchers.