A sensitive and fast diagnostic test for paratuberculosis is desirable. Cows from two known positive and three presumed negative herds were examined by cultivation of feces combined with PCR. The feces samples were collected at a field laboratory, decontaminated and cultivated on Löwenstein-Jensen (L-J) with mycobactin at once; then the samples were frozen at -20°C, thawed and recultivated. The feces samples were then sent to the Danish Veterinary Laboratory where they were decontaminated and cultivated in Dubos broth with mycobactin. The broths were analyzed after 1 day, 3 days, 1 week, 2 weeks, and 3 weeks by culture on L-J and by PCR. The L-J's were read once a week for 10 weeks. For PCR, the broths were spun down and the mycobacteria-pellets were lysed by bead beating with zirconium beads. DNA was amplified by PCR using primers specific for IS900. Samples with a low number of bacilli (<100 cfu/gm): Of 22 feces samples positive prior to freezing, only 6 were positive after freezing by primary cultivation on L-J (no pre-cultivation in Dubos), while 18 samples were positive after pre-cultivation in Dubos. Thirteen samples were found positive by PCR. One of these samples were culture negative. Samples with a high number of bacilli (>500 cfu/gm) were highly positive by cultivation prior to freezing. After freezing, these samples were positive in both culture (with and without pre-cultivation in Dubos) and by PCR. These results indicate that freezing of feces samples reduces the sensitivity of cultivation by approx. X10, while pre-cultivation in broth could enhance the sensitivity of cultivation. This combined with the improved PCR technique could considerably reduce the time needed for diagnosing paratuberculosis.