The MGIT 960 a high capacity, automated, non-invasive liquid culture system was proposed as an alternative to the BACTEC 460, a semi-automated, radiometric liquid culture system, for the growth of mycobacteria from bovine lymph node (LN) tissues. A limit of detection study for Mycobacterium bovis, spiked into negative LN tissue, showed that both liquid culture systems supported growth down to a concentration of 2.5 x 102 cfu/ml. A total of 867 routine LN tissues mainly from bovines but including some badger, deer and sheep tissues, were cultured on both liquid culture systems and on solid culture systems (Stonebrinks and LJP). For MTBC (Mycobacterium tuberculosis complex) isolates 92%, 93.2% and 73.9% grew on BACTEC 460, MGIT 960 and solid culture systems respectively. MGIT 960 detected growth of MTBC isolates slightly faster than the BACTEC 460 system (14.6 days compared to 15.2 days respectively). For Rhodococcus equi isolates 42%, 96.6% and 42% grew on BACTEC 460, MGIT 960 and solid culture systems respectively. For atypical mycobacteria isolates 89.5%, 43.4% and 10.5% grew on BACTEC 460, MGIT 960 and solid culture systems respectively. For Mycobacterium avium isolates 33.3%, 100% and 33.3% grew on BACTEC 460, MGIT 960 and Solid culture systems respectively. Contamination rate was 3.1%, 8.2% and 10.9% for BACTEC 460, MGIT 960 and solid culture systems respectively. For growth of MTBC, M. avium and R. equi isolates MGIT 960 system gave a similar or better detection rate compared to BACTEC 460 but gave a poorer isolation rate for the growth of atypical mycobacteria. Even with the higher contamination rate the MGIT 960 system was a suitable alternative to the BACTEC 460 system. A decontamination study, comparing 5% oxalic acid and 0.75% CPC (cetylpyridinium chloride), was carried out to determine if CPC could be used with the MGIT 960 system. CPC is not a suitable decontamination chemical for use with MGIT 960 system since only 47.4% of MTBC isolates were identified with CPC compared to 100% with oxalic acid decontamination. A study to compare five extraction methods to detect mycobacterial DNA directly from decontaminated LN tissue to determine which if any could be used, with culture, for rapid diagnosis of high priority samples such as singleton reactors and samples with potential human involvement. This study showed that heat lysis and the Roche High Pure Template Preparation kit were not specific or sensitive enough for this purpose. Immunomagnetic separation (IMS) gave a limit of detection result for M. bovis of 2.3 x 102 cfu/ml when spiked into PBS but this increased to 2.3 x 103 cfu/ml when spiked into negative LN tissue. IMS is not sensitive enough for the detection of mycobacterial DNA directly from decontaminated LN tissue. GenoType Mycobacteria Direct kit, with the addition of an overnight Proteinase K digestion, correctly identified 63.3% of the decontaminated LN tissues with 16.7% false negatives, the remainder of the samples failed. A persistent problem with the test was the failure of the positive control sample, supplied with the kit, to be correctly identified. Sequence Capture PCR, a known method for the detection of mycobacterial DNA in LN tissue, correctly identified 73.3% of the samples and the remaining 26.7% samples gave false negative results. Both the GenoType Mycobacteria Direct kit and Sequence Capture PCR could be suitable for the direct detection of mycobacterial DNA from decontaminated LN tissues with further work required to reduce the number sample failures and improve the sensitivity of both tests.