Development of novel diagnostics for tuberculosis has so far been governed by the clinical requirement of improving the detection of patients with paucibacillary forms of the disease. For this aim, serological assays have been evaluated using several antigens, but were found insufficiently sensitive, because antibody production associates with the bacterial load of the disease. Consequently, detection of antibodies against a relatively small number of selected well-defined antigens has a much higher sensitivity for sputum smear-positive pulmonary disease in adult HIV-negative patients. They are the most active in generating and spreading aerosols containing live tubercle bacilli, but their detection is often delayed, thus perpetuating the transmission of the infection and disease in the population. High volume throughput serological screening of clinical suspects with mild clinical symptoms may help to achieve diagnosis earlier, than currently used procedures. Such expanded testing could be done more efficiently in laboratories, than at ‘points-of-care’ and at a lower cost than other tests. The feasibility of this approach towards reducing the delayed diagnosis of the most infectious cases of pulmonary tuberculosis needs to be ascertained in prospective diagnostic trials, in populations at a high risk. Reducing the transmission of tuberculosis is of key importance for achieving its continued decline and therefore it is proposed, that the aims of serological screening should shift from clinical to public health priorities.