The main contribution of this work is a smart-charging strategy based on the state of charge analyses to avoid the undervoltages caused by plug-in electric vehicles into the distribution system. The work uses power hardware-in-the-loop simulations, where a modified IEEE 34 bus system, with five groups of electric vehicle stations, is modeled in a real-time digital simulator. The number of electric vehicles charging and the initial state of charge (SoC) are generated randomly. The first analyses identify the undervoltage problems due to the increased number of electric vehicles connected during peak hours load. After, a smart-charging solution is necessary to solve this power quality problem. So, if the voltage profile decreases under certain limits, defined by a hysteresis band, the control occurs, and the smart-charging is applied. The proposed strategy based on individual state of charge priority reduces the electric vehicle recharge power at virtual stations, by comparing its value with the mean of state of charge values from each station group. The experimental results presented show that, by varying the recharge current, the voltage profiles do not reach the limit for sags determined by the Brazilian standard, proving the performance and solving the problem even if not reducing the recharge current of all electric vehicles equally.