Mycobacterium avium subsp. paratuberculosis (MAP) is the cause of paratuberculosis or Johne's disease (JD) in cattle, leading to significant economic losses in the effected herds. Recently, boot swab samples have been used in cattle farms in order to establish the MAP-herd status. The aim of the present study was to evaluate the reproducibility of MAP-negative boot swab samples on cattle premises. Therefore, boot swab samples were collected from 158 farms with an average size of 16 cows, previously tested MAP-negative in boot swabs by the local veterinarians. Herds were selected based on the presence of risk factors for paratuberculosis (74 herds), sampling was performed by veterinarians who collected negative samples in all of their herds in a previous survey (75 herds) and clinical paratuberculosis cases (nine herds). Boot swabs were tested for MAP by solid culture and PCR, leading to positive results in eight herds (5.1%). Culture and PCR showed a high correlation, as seven of the herds were positive both in culture and PCR, and one herd was positive in PCR only. Based on these results the reproducibility of a negative boot swab result was estimated as 94.9% (95% CI: 90%, 98%). No significant differences in the results between veterinarians with and without positive samples in the first survey were found, indicating no or little influence of the person collecting the boot swab samples. From the results of the study it is concluded that negative boot swab sample results have a high reproducibility for the detection of MAP in cattle herds in small structured agricultural systems.