The BBL-mycobacteria growth indicator tube system (MGIT) is used for a rapid detection of the presence of mycobacteria. Our study aimed to compare MGIT with the Lowenstein-Jensen (LJ) reference method in clinical samples with suspected pulmonary and extrapulmonary tuberculosis, and to evaluate the primary and secondary resistance patterns by determining the resistances of the isolated strains to four major anti mycobacteria I drugs. 648 clinical samples from different clinics, with suspected pulmonary or extrapulmonary tuberculosis based on clinical, radiological, histopathological and immunological findings, were included in the investigation. The samples were first stained with Ziehl-Neelsen (ZN) and then cultured in LJ medium according to the standard bacteriological procedure and in the MGIT as recommended by the manufacturer. Conventional biochemical tests and p-nitro-alpha-acethylamino-beta-hydroxypropiophene of the Bactec system were used to identify the isolated mycobacterial strains. The susceptibilities to streptomycin, isoniazid, rifampicin, ethambutol were tested by the BBL-MGIT antibiotic susceptibility test and the resistances of the strains found to be resistant to any of the drugs were confirmed by the agar proportion method. Mycobacterium spp. were isolated in 61 (9.4%) out of 648 samples. Eventually, 58 out of 61 strains were classified as Mycobacterium tuberculosis and the other 3 as Mycobacterium tuberculosis complex. 32 of these were ZN positive. The growth time was determined as 12.2 days by the MGIT method and 24.1 days by the LJ method (p < 0.001). 29 strains were ZN negative. Their growth time was 23 days by the MGIT method and 37 days by the LJ method (p < 0.001). Drug resistance was detected in 23 (37.7%) of 61 cases (of whom 39 were new and 22 were former patients); of these resistances, 8 (20.51%) were primary and 15 (68.18%) were secondary. In double drug resistance, secondary resistance was found only to isoniazid + rifampin (4 cases) whereas both primary and secondary resistances were found to one drug. The highest cumulative drug resistance - both primary and secondary - was found to isoniazid. In conclusion, the MGIT was found to be advantageous because it enables rapid bacterial identification of tuberculosis and detection of antimicrobial resistance due to its high sensitivity and specificity. It is quicker than the LJ method. Its antibiotic susceptibility can be tested and it is easy to perform. We recommend to include it in routine laboratory work. In addition, our study suggests that the high ratio of secondary resistance in the public might be related to inappropriate and insufficient treatment of tuberculosis, and noncompliance, which appear to cause an important increase in primary tuberculosis as a result of new contaminations. Copyright (C) 2002 S. Karger AG, Basel.