The ultimate goal of evidence-based drug treatment is to produce a desired pharmacological response in a predictable manner and also to minimise adverse effects. This goal requires not only an increased awareness of the need to provide specific dosing recommendations aimed at specific patient groups, but also the implementation of a consistent integrative approach to recognise all factors contributing to the within- and between-subject variability in drug disposition and response. The assessment of new anti-tuberculosis agents and regimens in children requires a specific programme of investigation, and should be included early in human drug evaluation programmes. Appreciation of this principle is an important step forward towards the full integration of children into the tuberculosis research agenda and control programmes. The development of anti-tuberculosis drug formulations and regimens tailored to the requirements of children needs to consider physiological age-related differences for pharmacokinetics and toxicity between adults and children. Research based on these principles will create an evidence base that will inform the appropriate treatment of children with novel agents and regimens and will also inform future research, including the use of chemoprophylaxis and treatment-shortening strategies in children.