The prevalence of pulmonary infections due to non-tuberculous mycobacteria such as Mycobacterium abscessus, has been increasing and surpassing tuberculosis (TB) in some industrialized countries. Due to intrinsic resistance to most antibiotics that drastically limits conventional chemotherapeutic treatment options, new anti-M. abscessus therapeutics are urgently needed against this emerging pathogen. Extensive screening of a library of benzimidazole derivatives that were previously shown to be active against Mycobacterium tuberculosis led to the identification of a lead compound exhibiting very potent in vitro activity against a wide panel of M. abscessus clinical strains. Designated EJMCh-6, this compound, a 2-(2-cyclohexylethyl)-5,6-dimethyl-1H-benzo[d]imidazole), also exerted a very strong activity against intramacrophage-residing M. abscessus. Moreover, treatment of infected zebrafish embryos with EJMCh-6 correlated with a significantly increased embryo survival and a decrease in the bacterial burden as compared to untreated fish. Insights into the mechanism of action were inferred from the generation of spontaneous benzimidazole-resistant strains and the identification of a large set of missense mutations in MmpL3; the mycolic acid transporter in mycobacteria. Over-expression of the mutated mmpL3 alleles in a susceptible M. abscessus strain was associated with high resistance levels to EJMCh-6 and to other known MmpL3 inhibitors. Mapping the mutations conferring resistance on a MmpL3 three-dimensional homology model defined a potential EJMCh-6-binding cavity. These data emphasize a yet unexploited chemical structure class against M. abscessus with promising translational developments for the treatment of M. abscessus lung diseases.