Mycobacterium avium subspecies paratuberculosis (MAP) is the causative agent of paratuberculosis, which is currently prevalent in many parts of Iran and produces severe economic loss. It is hence necessary to identify and isolate the animals infected with this bacterium, so this research aimed to isolate MAP from milk and fecal samples of ELISA-positive animals and determine the molecular identity of isolates. After performing ELISA on 3,700 bovine blood samples, 115 samples of milk and feces were taken from ELISA-positive cattle and were cultured on Herald's egg yolk medium with and without mycobactin-J and then the acid-fastness of positive samples was determined using Ziehl-Neelsen staining. The 16S rRNA-PCR test was performed after DNA extraction to determine the molecular identity of isolates. Primers IS6110 and IS901 were employed to ensure that the isolates were not related to members of M. tuberculosis complex and M. avium, respectively. Primer IS900 was also used to determine the molecular identity of MAP isolates. Also, expression levels of MAP-related genes (IS900, ISMAP02, F57, MAP2191, MAP4027) were evaluated via qPCR. Finally, positive samples were confirmed based on the Nested-PCR. Results showed that a total of 9 isolates were obtained from the culture of 90 ELISA-positive samples. The results revealed that all grown samples were positive for acid-fastness. The 16S rRNA-PCR test revealed the 543 bp band, which confirms the presence of Mycobacterium in the samples. The PCR test with Primer IS900 generated the 398 bp fragment in the first step and the 298 bp fragment in the second step, indicating the presence of MAP in samples. Also, relative expression analysis revealed that MAP-related genes were significantly higher in ELIZA-positive samples than in negative ones. Based on the study findings, it can be concluded that MAP-infected animals can be identified by ELISA. In addition, mycobacterium can be isolated by culturing the samples on appropriate media and then its molecular identity can be determined by using nested-PCR.