Accurate online state of energy (SOE) estimation of a series-connected lithium-ion battery pack is very important for the driving range estimation of electric vehicles, which is still an urgent problem to be solved due to the high computational complexity caused by cell-to-cell inconsistency under practical application of the existing battery management system. In the manuscript, a novel low-complexity SOE simplified estimation method for lithium-ion battery pack based on prediction and representative cells is proposed. Firstly, two cells in the battery pack, one of which has the smallest remaining discharge capacity and the other of which reaches at or closest to the lower cut-off voltage during the discharge process, are selected as representative cells. Secondly, after the EKF algorithm is applied to achieve SOC estimation, the future states of the representative cells are accurately predicted to obtain the cut-off discharge time and the discharge capacity based on the thermoelectric coupling equivalent circuit model under the condition that the future output current of the battery pack has been predicted. The output energy of the representative cells during prediction is divided into three output energy generation parts contributed by OCV, ohmic resistance and polarization resistance respectively, and the three parts are cumulatively calculated and summed to obtain the remaining discharge energy of the representative cells. Thirdly, the approximately proportional relationships in three output energy generation parts between the representative and the non-representative cells are established based on the approximately proportional expressions of the three parts of all cells, and a simplified calculation method is subsequently applied to obtain the remaining discharge energy of non-representative cells based on the results of representative cells. Finally, the remaining discharge energy and SOE of the battery pack can be obtained. The results verified by complex dynamic condition experiments at different temperatures show that the maximum absolute error of the SOE estimation results is less than 3% and more than 85% of the calculation time is saved, which demonstrates that the developed method can achieve accurate and simple SOE estimation of the battery pack.