The objective of this study was to characterize the distribution of Mycobacterium avium subsp. paratuberculosis (Map) in the environment of infected and uninfected Minnesota dairy farms. One hundred and eight Minnesota dairy herds were sampled during the summer of 2002, including 80 herds known to be infected based on previous testing in the Johne's Disease Control Program (JDCP) of the Minnesota Board of Animal Health (MBAH) and 28 herds known to be uninfected based on previous testing in the Voluntary Johne's Disease Herd Status Program (VJDHSP) of the MBAH. Fecal samples were obtained from up to 100 cows in each herd and were cultured in pools of five cows per pool. Environmental samples were obtained from each farm, with up to 2 samples collected from various locations.Sixty-four of the 80 JDCP herds (80%) had at least one positive pool; 16 of these herds did not have any positive pools. The farm environment was determined to be contaminated on 61 of the 64 (95%) herds with at least one positive pool and in one of the 16 JDCP herds with no positive pools. Twenty-six of the 28 VJDHSP herds (93%) had no positive pools; 2 herds had one positive pool each. One environment sample from these herds was also cultured positive. The farm environment was determined to be positive in samples from cow alleyways (77% of the herds), manure storage (68%), calving area (21%), sick cow pen (18%), water stream edge and water runoff (6%), and postweaned calves area (3%), but not in preweaned calves or fields near cow area. There was an association between maximum level of colonies per tube from cow alleyways and manure storage and fecal pool prevalence. Herds with both areas cultured negative were estimated to have 0.3-4% fecal pool prevalence. Herds with both areas having a heavy load of bacteria were estimated to have 53-73% fecal pool prevalence. The study results indicate that targeted sampling of cow alleyways and manure storage may be a useful alternative strategy for herd screening and Johne's infection status assessment and for estimating herd fecal prevalence.