A cross-sectional study of bovine tuberculosis (BTB) was conducted in pastoral cattle herds in southern Ethiopia, from February to August 2008 using the comparative intradermal tuberculin test. The prevalence of BTB and the risk factors for having positive reactor herds were assessed in four pastoral associations in two districts of southern Ethiopia, namely Goro-Dola with 242 cattle in 16 herds and Liben with 231 cattle in 15 herds. A herd was considered positive if there was at least one reactor animal in a herd. The test results were interpreted based on the Office Internationale des Epizooties recommended 4-mm and a recently suggested 2-mm cut-off. The apparent individual animal prevalence of tuberculin reactors was 5.5% (95% confidence interval (CI), 4.0-8.0%) and 7.0% (95% CI, 5.0-10.0%), whereas the true prevalence estimate was 4.4% (95% CI, 0.8-8.0%) and 6.1% (95% CI, 2.6-9.5%), when using the 4-mm and the 2-mm cut-offs, respectively. The overall herd apparent prevalence of tuberculin reactor animals was 41.9% (95% CI, 24.9-60.9%) and 48.4% (95% CI, 30.2-66.9%) with the 4-mm and 2-mm cut-offs, respectively. A positive tuberculin test was associated with the age of animals and the main drinking water sources during dry seasons. In order to investigate the public health risks and the epidemiological importance of BTB in the area, we recommend to include other livestock species (camels and goats) as well as humans in future studies.