Seventy bovine fecal and serum samples were distributed to twenty laboratories for testing. Forty-six samples were obtained fro cattle with at least one previous positive fecal culture. The number of positive cultures in this group of samples from any one laboratory varied from 3 to 24. Thirty-two (70%) of the infected animals were correctly classified by fecal culture in at least one laboratory. Only five of the 1400 fecal samples tested by the twenty laboratories were falsely classified as positive. The DNA probe was positive on six of the 70 fecal samples tested. A commercial test for Johne's disease antibody was positive for three of the 70 sera tested. Several laboratories ran the complement fixation test under different protocols, with variable results. Although culture protocols have been published, many diagnostic laboratories need to update the methods used to detect M. paratuberculosis in fecal samples to improve test sensitivity.