Serology is currently used as the primary screening test for flock diagnosis of ovine Johne's disease in New South Wales. We have previously shown that in flocks where there is a high prevalence of disease (>6%), detection of infection can be significantly enhanced by biasing the test sample to include sheep of low body weight or condition score. This study reports the association between low bodyweight and condition score and the presence of Johne's disease in flocks where there is a low prevalence (<5%) of disease. We sampled 1,300 sheep from 5 flocks known to be infected with Johne's flocks. These flocks were derived from 4 different properties. All sheep were individually identified by ear tag and their bodyweight and condition score recorded. Serum collected from each sheep was tested with AGID and absorbed ELISA techniques, and a 5cm portion of terminal ileum collected from each sheep at slaughter was processed for histological examination. The prevalence of Johne's disease in each flock was low and no greater than 5% depending on the diagnostic method used. Use of biased sampling techniques based on low bodyweight and condition score did not always enhance detection of infection in the flocks studied. Further work is necessary to develop the best strategy for detection of Johne's disease in flocks were there is likely to be a low prevalence of disease.