Given the current need for predictive persisting model for , we adopted a classical assay to study drug-tolerant bacterial persisters, focusing on the behavior of a small antibiotic-insensitive subpopulation during prolonged exposure to moxifloxacin. Our study showed a wide-ranging response of , depending on antibiotic concentration, growth stage of mycobacterial cultures, and the availability of potassium ions in the medium. Mid-logarithmic cultures, initially grown in either balanced or K-free medium, contained small sup-populations capable of prolonged and stable survival in the presence of moxifloxacin. The response of these mid-log cultures to antibiotic exposure involved initial killing, followed by regrowth at 1-2 MBCs of moxifloxacin or a substantial reduction of the antibiotic-insensitive subpopulation to fewer than 10 CFU/mL at 16 MBCs. In stationary-phase cultures grown in a complete medium, a consistent number of viable cells was observed when exposed to a high dose of moxifloxacin. In contrast, antibiotic-insensitive subpopulations in stationary-phase cultures under potassium-deficient conditions experienced gradual killing across a wide range of moxifloxacin concentrations (1-16 MBCs). Studies on electron microscopy demonstrated that singular cells were rapidly destroyed after relatively short-term exposure to moxifloxacin, while cells in aggregates or clumps persisted longer, explaining the delayed biocidal effect. The small subpopulation that survived under intense moxifloxacin pressure was notably heterogeneous in cell morphology and fine structure, consisting of ovoid forms and cell-wall-deficient cells with reduced size. These findings suggest that the same antibiotic dose may have varying effects on cells, depending on their physiological state and abundance within infected cells or tissues. Taken together, our study may contribute to the development of strategies to combat recalcitrant survivor subpopulations.