We investigated the prevalence and risk factors to positive herd-level tuberculin reactivity between October 2003 to May 2004 to bovine tuberculosis (BTB) in the four transhumant districts of Uganda: three districts (Karamoja region) of nomadic transhumance cattle rearing (30 superherds and 1522 cattle), and one district (Nakasongola) of fixed-transhumance (7 herds and 342 cattle). We used the comparative intradermal skin-test, sampled 50 animals per superherd/herd, and considered herd positive if there was at least one reactor. Of the 30 superherds under nomadic transhumance, 60% (95% CI 41.4, 79) were tuberculin-test positive; of the 7 fixed herds, 14.3% (95% CI -20.7, 49.2) were tuberculin test positive. The true herd prevalence was estimated at 46.6%. Many risk factors were collinear. The final multivariable logistic-regression model included: recent introductions from market (OR=3.4; 95% CI 1.1, 10.3), drinking water form mud holes during dry season (OR=49; 95% CI 9.1, 262), and the presence of monkeys (OR=0.08; 95% CI 0.0, 0.6) or warthogs (OR=0.1; 95% CI 0.0, 0.3). No association was found between herd size or number of herd contacts with reactors; it was probably masked by the effect of high between-herd interactions. Provision of water from mud holes in dry river beds and introductions of new animals are risk factors that might be targeted to control BTB in transhumance areas.