We started out spiking milk samples from cows tested negative for paratuberculosis. This was in order to establish the PCR and cultivation procedures and also to investigate in which milk fractions (cream, pellet or whey) M. paratuberculosis was found in largest numbers. Our results showed bacteria in pellet and cream but hardly any in whey. After this we sent out tubes for milk sampling to practitioners if faeces samples from their patients had shown the prescence of acid fast bacilli. The practitioner was asked to take four milk samples (one from each quarter after a proper outer disinfection of the udder) from each cow. When we got the tubes back we spun down the milk and performed culture (directly on solid medium and in Dubos broth for 2 weeks followed by culture on solid medium) and PCR on both the pellet and the cream fractions. Of 34 sets of tubes sent out, we received 12 sets back at the laboratory. M. paratuberculosis was cultivated in innumerable numbers from 8 of these 12 animals faeces or intestinal mucosa. From 5 cows milk (all faeces culture positive) we cultivated M. paratuberculosis in few numbers (less than 10 colonies pr. tube) and milk from 2 cows were PCR positive (both animals were faeces culture positive and 1 cow was milk culture positive). One cow was culture negative on intestinal mucosa but culture positive in milk and 3 cows were negative in culture and PCR from both faeces and milk. M. paratuberculosis was equally found in pellet or cream. In conclusion: We can show the presence of M. paratuberculosis in milk by PCR but cultivation of milk is more sensitive.