Young children are most vulnerable to develop severe forms of tuberculosis (TB) and are overrepresented among TB deaths. Almost all children estimated to have died from TB were never diagnosed or offered TB treatment. Improved access to TB preventive treatment (TPT) requires major upscaling of household contact investigation with allocation of adequate resources. Symptom-based screening is often discouraged in adults for fear of generating drug resistance, if TB cases are missed. However, the situation in vulnerable young children is different, as they present minimal risk of drug resistance generation. Further, the perceived need for additional diagnostic evaluation presents a major barrier to TPT access and underlies general reluctance to consider pragmatic decentralised models of care. Widespread roll-out of Xpert MTB/RIF Ultra 1 represents an opportunity for improved case detection in young children, but attaining full impact will require the use of non-sputum specimens. The new Fujifilm SILVAMP TB LAM 1 urine assay demonstrated good diagnostic accuracy in HIV-positive and malnourished children, but further validation is required. Given the limited accuracy of all available tests and the excellent tolerance of TB drugs in children, the global community may have to accept some overtreatment if we want to close the persistent case detection gap in young children.