A national tuberculin testing trial was carried out in 8151 cattle in 132 herds due for normal herd testing and distributed throughout the 11 animal health administrative regions of England, Scotland and Wales. The numbers selected from each region were in approximately equal proportions to the total cattle populations of the regions and comprised a representative sample of the national herd. The standard intradermal comparative tuberculin test using Weybridge avian and human purified protein derivatives (PPD) was carried out on one side of the neck, and Rotterdam bovine PPD (1 X 0 mg per ml) was injected on the opposite side in the usual 'mammalian' site. Of the cattle tested, 9 X 4 per cent were classified as positive reactors to a single intradermal injection of the bovine tuberculin applying the official EEC interpretation for tuberculin tests. A single intradermal injection of human PPD would have resulted in a significantly higher proportion (10 X 4 per cent) of cattle being designated for removal. The rules of interpretation developed by Lesslie and Hebert (1975) for a comparative tuberculin test using avian and bovine PPDs, when applied to this sample, classified 0 X 80 per cent as inconclusive reactors compared with 0 X 58 per cent for the avian/human comparative test. The avian/bovine test also classified as positive reactors 0 X 12 per cent of the cattle tested, whilst the avian/human test classified as positive 0 X 05 per cent of cattle in routine herd tests during the three-year period 1972 to 1974.