Moxifloxacin (MOX), an 8-methoxyquinolone compound, is now widely used for the treatment of bacterial infections and also accepted as 2nd-line drug for the treatment of multidrug-resistant (MDR) tuberculosis. To tentatively correlate the clinical outcome with in vitro results, we sought to set up susceptibility test conditions for Mycobacterium tuberculosis against MOX by using the reference agar proportion method, the BACTEC 460 radiometric system, and the recently validated nonradiometric BACTEC MGIT 960 system. Our aim was to determine the critical MOX test concentration to be used with the abovementioned methods for routine susceptibility testing. MICs were determined for 20 pan-susceptible strains, 10 MDR strains, and 10 fluoroquinolone-resistant strains with defined gyrA mutations. MOX MICs resulted in a bimodal pattern with values for gyrA mutants considerably higher than those for pan-susceptible and MDR strains. Our data showed that a concentration of 0.5 microg/mL allowed a clear-cut separation between susceptible and resistant strains when tested by all the studied methods. Confirmatory test with a subset of pan-susceptible and MDR isolates appeared to validate the selected critical concentration. The MOX-resistant strains were almost isolated from patients with prior fluoroquinolone exposure.