Tuberculosis (TB) in cattle remains a major zoonotic and economic problem in many countries. Since the standard diagnostic assay, the intradermal test (IDT) with bovine PPD tuberculin, has less than optimal accuracy in all situations, other diagnostic methods such as serological assays have been investigated. Because of fundamental concerns for the low sensitivity and specificity of previous ELISA protocols, a profiling ELISA with nine purified, recombinant proteins of TB complex mycobacteria, was employed on samples from four groups of cattle: (a) naturally Mycobacterium avium-exposed and experimentally Mycobacterium bovis-infected, (b) officially-certified TB-free herds, (c) exposed to M. bovis in two field TB outbreaks and scored as bovine reactors in the gamma-IFN assay for bovine TB, (d) paratuberculosis (para TB)-infected. The described ELISA proved to be highly specific. In fact, the antibody (Ab) response could be consistently detected in 3 out of 3 endotracheally-infected calves and in 1 out of 3 contact-infected calves. There was also a very low prevalence of low-titered, non-specific Ab responses in paraTB-infected animals. As for the animals exposed to field TB outbreaks, 16 out of 28 gamma-IFN positive cattle were also Ab-positive; importantly, 7 out of 12 gamma-IFN positive, IDT-negative cattle showed Ab responses to TB proteins. In general, the profile of the Ab response varied among animals; the reaction to single recombinant antigens was sometimes transient and fluctuating, whereas the panel of antigens on the whole was indeed more effective in Ab detection.