Host immune responses to Mycobacterium bovis (M. bovis) infection are variable at the different severity stages of pathology of the disease. In countries like Ethiopia, where routine screening of bovine TB is not undertaken, the use of tests which measure cellular and antibody responses may help for the maximum detection of infection. In the present study, 701 cattle were tested for bovine tuberculosis (BTB) using comparative intradermal tuberculin (CIDT) test, interferon (IFN)-gamma test, and lateral flow assay. The apparent prevalence was 32% when all the three tests were used, but varied from 23 to 25% when a pair of tests was used and from 9% to 15% when a single test was used. Agreement was observed between CIDT and IFN-gamma tests both at a cut-off >2 mm (Kappa +/- standard Error, k +/- SE, 0.129 +/- 0.045; 95%CI = 0.041,0.216) and a cut-off >4 mm (k +/- SE, 0.094 +/- 0.044, 95%CI = 0.008,0.179) while no agreement was observed either between CIDT test and lateral flow assay (k +/- SE, -0.04 +/- 0.033; 95%CI = -0.104,0.024) or between IFN-gamma tests and lateral flow assay (k +/- SE, -0.031 +/- 0.032; 95% CI = -0.093,0.031). Thus, the use of more than one test leads to the detection of the maximum number of infected animals.