Estimated herd prevalence of bovine paratuberculosis in Germany is 10 to 15%, resulting in high economic losses for dairy farmers. The possible link between Mycobacterium avium subspecies paratuberculosis (MAP) and Morbus Crohn has not been elucidated to date, this increases the need for a control and surveillance program. Laboratory diagnosis of paratuberculosis is labor intensive, complicated and time consuming and extensive, area-wide testing with these methods would not be possible. An alternative and affordable option would be a two-step testing of bulk milk to estimate herd prevalence. The first step determines the antibody titer in the bulk milk with a highly sensitive test (Svanovir®), the second step includes detection of MAP in bulk milk from “non negative” herds using a capture-PCR. Using this approach, our results indicate a correlation of milk ELISA and MAP detection via capture-PCR, as we predominantly could detect MAP-DNA in milk which was serologically tested as “recommended for surveillance” (PP value 5-15) or “highly prevalent” (PP value >16). Based on our findings, the suggested combination of serological and molecular biological methods seems to be reasonable for herd classification. The combination of the test systems facilitates definition of the herd status and gives the opportunity to protect the consumer by excluding MAP containing milk from the dairy manufacturing process. If MAP is detected in milk, testing on single animal level becomes necessary to exclude shedders from the herd.