A tuberculin testing trial in cattle was carried out in the Republic of Ireland to compare the specificity for bovine tuberculosis of a human purified protein derivative (PPD) tuberculin (Weybridge) with that of a bovine PPD (Rotterdam), and to determine whether discrimination between specific and non-specific reactions to mammalian tuberculin is better with doses of tuberculins smaller than those traditonally used for testing cattle. Tests were carried out in 510 cattle, 395 of which were shown by post mortem examination to be tuberculous and 115 non-tuberculous. Three dilutions at five-fold intervals of both mammalian tuberculins were used together with two dose levels of avian tuberculin PPD (Weybridge), and all reactions were measured both by increase in skin fold thickness and by diameter of induration. In the environment of this trial, the bovine PPD was shown to be more specific for bovine tuberculosis than the human PPD, and particularly in differentiating from "skin tuberculosis". There was no indication of greater specificity at lower doses of tuberculin. Measurement of induration diameter proved a satisfactory alternative method of reading tuberculin reactions in cattle under field conditions.