Rapid and accurate diagnosis of mycobacteria is very important in the prevention and effective treatment of tuberculosis which is still a serious public health problem. Fluorescence in situ hybridization (FISH) method using rRNA targeted probes allows for precise and accurate identification of mixed microorganisms from cultures and directly from clinical samples within a few hours without the need for culture methods. In this study it was aimed to compare the diagnostic performance of two different FISH methods (Oligo-FISH and PNA-FISH) with the conventional culture methods for the identification of Mycobacterium spp. grown in BACTEC MGIT (TM) (Mycobacteria Growth Indicator Tube) system. A total of 60 MGIT (BD, USA) positive, 52 MGIT negative samples and 10 different reference strains were included in the study. 16S rRNA targeted oligonucleotide probes (Myc657: Mycobacterium subdivision, Eub338: Positive control, NonEub: Negative control) were used for oligo-FISH, and 16S rRNA targeted peptide nucleotide probes (MTC: Mycobacterium tuberculosis complex, NTM: Non-tuberculosis Mycobacterium, BacUni: Positive control) for PNA-FISH. Ehrlich-Ziehl-Neelsen staining (ARB) and Lowenstein-Jensen (LJ) culture methods were performed as conventional methods as well as MGIT 960 culture system. Of MGIT positive 60 samples (44 sputum, 4 tissue, 4 urine, 3 bronchoalveolar lavage, 3 CSF, 1 abscess, 1 peritoneal fluid), 29 (48.3%) were found positive for ARB and 44 (73.3%) with LJ culture methods giving a total of 59 positive results. Fifty-eight (96.6%) of those isolates were identified as MTC, and one (1.7%) as NTM by conventional methods. By using Oligo-FISH, 95% (57/60) of the isolates were identified as Mycobacterium spp., while three samples (5%) yielded negative result. By using PNA-FISH, 54 (91.5%) isolates were identified as mycobacteria, of them 53 (90%) were typed as MTC and 1 (1.7%) as NTM. Five isolates that were found positive with Oligo-FISH, but negative with PNA-FISH, yielded positive result with PNA-FISH method performed with minor modifications. It was determined that both FISH methods are more rapid (approximately 2-2.5 hours) and practical than the conventional culture methods and also PNA-FISH was more practical than Oligo-FISH. The sensitivity, specificity, positive and negative predictive values of the probes used for Oligo-FISH, were 96.6%, 100%, 100% and 96.4%, respectively. Those values for the probes used for PNA-FISH, were 91.5%, 100%, 100% and 91.4%, respectively (p< 0.0001). The compatibility of the methods was calculated with kappa statistical analysis, assigning perfect concordances between Oligo- and PNA-FISH methods, as well as between conventional and both of the FISH methods (kappa: 0.964, 0.929, 0.964; p= 0.001). The coverage of oligonucleotide and PNA probes was also checked by using 16S rRNA gene sequence database retrieved from the SILVA 102. It was determined that the rates of coverage were 86.5% for Eub338, 41.7% for Myc657, 84.2% for BacUni, 76.3% for MTC (100% for only M.tuberculosis and M.bovis) and 25.8% for NTM probes. In conclusion, Oligo- and PNA-FISH methods seem to be successful for rapid and accurate identification of Mycobacterium spp. from MGIT positive cultures in routine mycobacteriology laboratories without the need for expensive methods.