Capacitive deionization (CDI) has emerged as an efficient process for hardness control than traditional water-softening technologies. However, few systematic studies have been conducted on water softening using pilot-scale CDI systems. For industrial applications, it is essential to investigate deionization performance, energy consumption, and selectivity for divalent cations through pilot-scale studies. In this study, we examined the deionization performance and energy consumption of a pilot-scale CDI process for effective water softening. The key operational variables analyzed were the applied voltage (200–300 V (i.e., 1.0–1.5 V per pair)), feed concentration (550–1250 mg/L), recovery ratio (50–80 %), flow rate (3–5 L/min), and adsorption/desorption time (80–180 s). Within this range, a high voltage, low feed concentration, low recovery ratio, high flow rate, and long adsorption/desorption times are advantageous for the total dissolved solids (TDS) removal ratio. In terms of energy efficiency, expressed as TDS removal per unit energy consumption, a low voltage, low feed concentration, low recovery ratio, fast flow rate, and long adsorption/desorption time are beneficial. Additionally, we investigated the selective removal of Ca2+ over Na+ under various operating conditions, as well as its impact on enhancing energy efficiency. The pilot CDI system achieved a Ca/Na selectivity coefficient of approximately 2.4 and it was analyzed to achieve up to 45 % energy-saving efficiency compared to nonselective systems (i.e., selectivity coefficient = 1). The results of this study are expected to provide valuable data for the application of CDI to water softening and hardness control processes.