A retrospective study of various diagnostic postmortem techniques used in a 4-year surveillance program for detection of Mycobacterium bovis infection in wild white-tailed deer (Odocoileus virginianus) was conducted. The tests evaluated were routine histopathology, acid-fast staining, detection of acid-fast bacilli in culture, and an M. tuberculosis group-specific genetic probe applied to pure cultures. Each of these techniques were compared with a reference or "gold standard" of mycobacterial culture and identification. Histopathology, the most rapid form of testing for M. bovis infection in white-tailed deer samples, had a sensitivity of 98% and a specificity of 87%, resulting in a positive predictive value of 94%. The detection of acid-fast bacilli by staining was less sensitive than histopathology (90%), but its higher specificity (97%) resulted in a positive predictive value of 99%. The detection of acid-fast bacilli on culture was both highly specific (93%) and sensitive (100%). The group-specific genetic probe had the highest sensitivity and specificity and produced results in complete agreement with those of mycobacterial culture, suggesting that this technique could be used as the new "gold standard" for this particular wildlife tuberculosis surveillance program.