Most US bull studs have herd surveillance programs for Johne's disease (JD) to meet semen export requirements. However, due to lack of sensitivity of tests to detect early infection states and as JD becomes more prevalent, the risk of JD entering bull studs through AI source herds also increases. Because young AI bulls in waiting are typically group housed, the risk of horizontal infection with JD becomes an additional concern. Horizontal spread of JD among young bulls housed in the same pen was demonstrated at a large US Artificial Insemination facility. The primary semen production herd has been paratuberculosis test-negative since the 1970s. In April 1991, a 24 month old bull was detected on annual fecal culture to be shedding Mycobacterium paratuberculosis. The bull was group housed with eight other bulls of the same age for the previous 10 months. The infected bull was culture positive (too numerous to count (greater than 100) colony forming units (CFU)) on tissues taken at slaughter in June 1991 including ileum, ileocecal lymph nodes, seminal vesicles, and epididymis. The eight penmates were culture negative on bimonthly fecal samples from March 1991 to September 1992. However, on samples taken November 3, 1992, five of the eight were detected as shedding low (less than 10 CFU per 2 gm) numbers of M. paratuberculosis. From March 1993 to present (January 1994), weekly cultures on all eight bulls have remained negative. Serum has been negative ELISA (IDEXX Corp. Portland, ME) and low to moderate risk on KELA ELISA, developed in our lab and used as a herd screening test on the NYS Paratuberculosis Program. Infection in exposed older animals has been shown to occur, but little is known about the progression of infection of shedding or clinical disease, or the influence of health, management or genetic factors on the infection state. This natural experiment provided biologic and epidemiologic evidence that at least five of eight young bulls were horizontally infected with JD. All eight were exposed to the manure of an infected penmate from 16 to 27 months of age, and five were detected shedding M. paratuberculosis 19 months later. Persistently negative status on subsequent tests for all eight, however, questions the state or existence of the infection. These bulls are five years of age in 1994 and are being further monitored to answer that question. Horizontal spread of JD infection can occur in group housing conditions. When valuable animals and market are at risk, increased monitoring with multiple tests, heightened management and promoting to the industry the benefits of preventing JD are justified.