OBJECTIVE: In this retrospective study, the evaluation of M. tuberculosis complex rates and method-based differences were used to determine the resistance rates of culture positive samples to anti-tuberculosis drugs. MATERIAL AND METHODS: Six thousand, eight hundred and fourteen tuberculosis suspicious clinical samples were examined retrospectively in the Baskent University Faculty of Medicine Adana Hospital microbiology laboratory. After the NaOH-NALC method, all specimens were examined directly, cultured on Löwenstein-Jensen (L-J) and BACTEC 12B medium. Microscopic examinations for acid-fast bacillus were done with Ehrlich Ziehl Neelsen stain between 2005 and 2008 and auramine-rhodamine stain between 2008 and 2010. 373 samples identified as M. tuberculosis as a consequence of NAP testing using the BACTEC 460 system were tested for sensitivity to anti-tuberculosis drugs with a BACTEC 460 TB radiometric method. RESULTS: Sputum (2628) was the most common specimen, and biopsy samples had the highest positivity rates (18.5%). The sensitivity of acid-fast stain to observe tuberculosis rates was 50% and specificity was 99.7%; the BACTEC 460TB assay was considered the gold standard. The Löwenstein-Jensen culture method had 80% sensitivity and 99.7% specificity when compared with the BACTEC 460TB method. Single drug resistance rates in first line anti-tuberculosis drugs for ethambutol, INH, streptomycin, and rifampin were 3.2% (n=12), 2.9% (n=11), 2.1% (n=8), 0.5% (n=2), respectively. Multidrug resistance was determined in 8 strains (2.1%). CONCLUSION: The tuberculosis detection rate was 5.7% for all samples. Although multidrug resistance rates are not very high in our patient population, determination of resistance is important for treatment follow-up.