Herein, we demonstrated the suitability and effectiveness of utilizing flow-electrode capacitive deionization (FCDI) for treatment of fluoride-contaminated brackish groundwater. By comparing operational modes of short-circuited closed-cycle (SCC), isolated closed-cycle (ICC) and single cycle (SC), it was found that SCC mode was the most advantageous. In SCC configuration, the effects of different parameters on the removal of F^– and Cl^– were investigated including current density, hydraulic residence time (HRT), activated carbon (AC) loading and feed concentration of coexisting NaCl. Results indicated that the steady-state effluent Cl^– concentration dropped with elevated applied current, and the decreasing rate got faster with the increase of HRT or AC loading. However, for the steady-state effluent F^– concentration, it dropped to a value under a small applied current and maintained stable in spite of the increase in applied current, and both HRT and AC loading had insignificant effects on the steady-state effluent F^– concentration. F^– was preferentially removed from the treated water compared with Cl^–, and a higher ion selectivity could be obtained at lower applied current and smaller HRT with the trade-off being that operation under these circumstances would generate outlet water with little change in conductivity compared to the influent. The removal efficiencies of F^– and Cl^– both decreased with increasing feed concentration of coexisting NaCl. This study should be of value in establishing FCDI as a viable technology for treatment of fluoride-contaminated brackish groundwater.