We examined whether the BACTEC/Mycobacteria Growth Indicator Tube (MGIT) system alone could supplant the use of a supplemental Löwenstein-Jensen (LJ) slant for routine recovery of Mycobacterium species from clinical specimens. A total of 6,062 specimens were included in the study. Of these, 273 specimens were positive for 278 mycobacterial isolates while 15 specimens were smear positive but culture negative using both media. Further analysis showed that 143 (51.4%) of the 278 total isolates were recovered from both the MGIT and LJ media. An additional 106 isolates (38.1%) were recovered from the MGIT only, while 29 (10.4%) isolates grew only on the LJ slant. The overall sensitivities of the MGIT and LJ media were 86.5% and 59.7%, respectively, for the recovery of mycobacteria from clinical materials. This study shows that although the MGIT system demonstrates better sensitivity for the recovery of mycobacteria from clinical specimens, both media types are necessary to maximize the sensitivity of detection.