This work was conducted to evaluate the methods frequently used for the diagnosis of Tuberculosis (TB). The study was run on 175 subjects in two groups including 150 cases suspected to have urinary TB (patients group) and 25 healthy subjects (control group). The following investigations were performed: Z-N staining smears, culture on L-J medium, Amplicor-PCR for urine specimens and detection of IgM antibodies against M. tuberculosis in serum samples. Considering the culture results as the golden standard for diagnosis of TB, the sensitivity, specificity, positive and negative predictive values and accuracy for (i) Ziehl-Neelsen (Z-N) staining smears were 71.4, 31.8, 57.1, 46.7 and 54%, respectively, (ii) Amplicor- PCR were 83.3, 72.7, 79.5, 77.4 and 78.7%, respectively, and (iii) M tuberculosis-IgM antibodies were 27.4, 77.3, 60.5, 45.5 and 49.3%, respectively. On taking the consensus analysis as the golden standard, the sensitivity, specificity, positive and negative predictive values and accuracy for (i) Z-N staining smears were 75.8, 40, 68.6, 48.9 and 62.7%, respectively, (ii) Lowenstein-Jensen (L-J) culture were 85.3, 94.5, 96.4, 78.8 and 88.7%, respectively, (iii) Amplicor-PCR were 86.3, 89.1, 93.2, 79 and 87.3%, respectively, and (iv) M. tuberculosis-IgM antibodies were 31.6, 85.5, 78.9, 42 and 51.3%, respectively. It was concluded that (i) PCR is more rapid, sensitive, and specific than the currently available techniques, but more expensive, (ii) PCR can distinguish cases with positive smears or cultures due to atypical mycobacteria from tuberculous cases, (iii) PCR is a useful and rapid method in monitoring the efficacy of anti-TB treatment. Thus, PCR may be suggested to be the golden standard for the diagnosis of TB.