The sensitivity of the abattoir inspection procedure introduced for Australian export beef in 1976 was compared to a detailed necropsy procedure for the detection of tuberculous lesions in cattle. In a sample of cattle that were reactors to the tuberculin test, abattoir inspection failed to detect an estimated 47% of cattle with lesions. The detailed necropsy examination of cattle with lesions of tuberculosis identified 21 sites of infection compared with 13 to 18 in cattle examined by routine meat inspection procedures. Of the lesions detected during detailed necropsy, 15.9% did not involve the thoracic cavity or the medial retropharyngeal lymph nodes. The failure to detect lesions during abattoir inspection has its greatest significance in an animal with a single lesion. If the 245 cattle found with single lesions during detailed necropsy had been examined by abattoir inspection using the 1976 or the 1986 procedures, 0.8 and 8.9%, respectively, of these animals would not have been detected because the diseased tissues would not have been examined. If meat inspection is to provide an effective means of monitoring the level of bovine tuberculosis during the final stages of eradication, a procedure no less sensitive than that introduced in 1976 should be used.