A review of the literature on the history of the tuberculin test and causes of non-specific reactions in cattle is given. The economic importance of the nonspecific reactor problem in South Africa is discussed. The cultural and biochemical characteristics and in some instances the virulence for laboratory animals of the 42 strains used in the investigation are given. The strains used were typical of the species they represented in these characteristics, except for the scotochromogen Hg 3. Details are given of the preparation of PPD sensitins from one strain of M. hovis, three strains of M. kansasii, two strains of M. fortuitum, two strains of M. phlei, two strains of M. smegmatis, six M. avium and avian-like strains, two strains of scotochromogenic mycobacteria, and six strains of unclassified mycobacteria probably belonging to a single species. The allergenic characteristics of the 42 strains of Mycobacteria and the twenty four sensitins were studied in guinea pigs. Specificity differences of the sensitins and sensitivity profiles of the strains are given. The sensitivity caused by various species of mycobacteria was found to be species specific, although varying degrees of cross sensitivity do occur between different species. Guine1 pigs and cattle sensitized by all the species investigated except M. bovus is were more sensitive to avian PPD than to bovine PPD. It was shown that the specificity of the sensitivity caused by different species of mycobacteria is similar in calves and guinea pigs .Multiple comparative tests in naturally sensitized non-specific reactor cattle showed most of these animals to be more sensitive to avian tuberculin than to any of the other sensitins used in the investigation. The sensitivity profiles in these animals were similar to sensitivity profiles in guinea pigs sensitized by M. avium .This indicates that the most common cause of non-specific reactors in these herd was M. avium or avian-like mycobacteria .Field trials with Onderstepoort PPD tuberculins showed the comparative test to be more accurate than a test with bovine tuberculin alone in differentiating nonspecific reactors and tuberculous animals. The simultaneous injection of via and bovine tuberculin did not influence the sensitivity to the latter. Results are given of investigations in two herds where both non-specific reactors and tuberculous animals occurred. Non-specific reactions were more common in young and tuberculosis more common in older cattle. The most suitable interpretation standards for the comparative test and the test with bovine tuberculin alone were investigated. Suggested interpretation keys are given.