Objective:The aim of this study was to evaluate the role of polymerase chain reaction (PCR) in the diagnosis of spinal tuberculosis after 2 weeks of preoperative anti-tuberculosis treatment and to compare PCR to the Löwenstein -Jensen Culture (LJC) and histopathological examination (HPE) methods. Methods: Twenty-five patients were included in this study. Sixteen patients were diagnosed and treated for spinal tuberculosis based on clinical and radiological evidence. Nine patients were controls. The LJC method and HPE of the specimen were performed according to hospital protocol. PCR was performed using primer encoding insertion of sequences IS6110 for mycobacterium tuberculosis complex. Clinical findings and radiological features were the gold standard for comparison. Results: PCR results were 15 positive and one negative. The sensitivity and specificity of PCR was 94% and 100% respectively (with 95% confidence interval [CI] 67% to 99% and 63% to 100%, respectively). HPE results showed 13 were positive and 3 negative in the spinal tuberculosis group; for the control group, all were negative. Sensitivity and specificity value of HPE was 82 % and 100% respectively (with 95% confidence interval [CI] 54% to 95% and 63% to 100%, respectively). Use of LJC showed only one was positive and 15 were negative in the spinal tuberculosis group whole all nine in the control group were negative. Sensitivity and specificity value of LJC was 6% and 100% respectively (with 95% confidence interval [CI] 0.3% to 32% and 63% to 100%, respectively). Conclusion: Our findings showed that the PCR for Mycobacterium tuberculosis is reliable as a method for diagnosis of spinal tuberculosis, even after of 2 weeks of anti-TB treatment, with an overall sensitivity of 94% and specificity of 100%.