Indirect fluorescent antibody (IFA) tests were carried out on a total of 545 individual serums from beef cattle just before tuberculin testing, using specific FITC-anti-bovine IgG and IgM conjugates. Phenol-killed M. tuberculosis (H37Rv strain) or field strains of M. bovis and M. avium were used as antigenic substrates. All cattle were examined at abattoirs for the presence of tuberculous lesions. A statistically significant difference (p less than 0.001) was found between the number of cattle with serums containing IgM and those with IgG antibody to M. bovis. The majority of cross-reactions that occurred with M. avium was due to IgM antibody. The IFA test using M. bovis and both antiglobulin conjugates detected 77% of cattle with visible lesions of tuberculosis, while the single intradermal caudal fold tuberculin test detected 83%. The IFA test detected 25 of 31 cattle with lesions that failed to react in the tuberculin test. The tuberculin test gave apparently false positive reactions in 6.2% and the IFA test gave false positive reactions in 21.2% of all cattle. The IFA test is of potential use as a further aid in the detection of tuberculous cattle that fail to react to tuberculin. It offers no advantage over the tuberculin test for the detection of all forms of the disease because it fails to detect a significant proportion of tuberculous cattle and produces an unacceptably high number of false positive results.