The most widely used ante-mortem diagnostic tests for tuberculosis in cattle are the tuberculin skin test and the interferon-gamma (IFN-γ) release assay, both of which measure cell-mediated immune responses to Mycobacterium bovis infection. However, limitations in the performance of these tests results in a failure to identify all infected animals. In attempting to increase the range of diagnostic tests for tuberculosis, measurement of the cytokine IP-10 in antigen-stimulated blood has previously been shown to improve the detection of M. tuberculosis and M. bovis infection, in humans and African buffaloes (Syncerus caffer), respectively. In the present study, 60 cattle were identified by the single intradermal comparative tuberculin test as tuberculosis reactors (n = 24) or non-reactors (n = 36) and the release of IFN-γ and IP-10 in antigen-stimulated whole blood from these animals was measured using bovine specific ELISAs. There was a strong correlation between IP-10 and IFN-γ production in these samples. Moreover, measurement of the differential release of IP-10 in response to stimulation with M. bovis purified protein derivative (PPD) and M. avium PPD distinguished between reactor and non-reactor cattle with a sensitivity of 100% (95% CI, 86%-100%) and a specificity of 97% (95% CI, 85%-100%). These results suggest that IP-10 might prove valuable as a diagnostic biomarker of M. bovis infection in cattle.