Lithium-ion battery packs require the construction of health index to indirectly represent the state of degradation. The health index construction method based on voltage curve suffers from the difficulty of parameter selection and the influence of health index representation ability by working conditions. To address the above problems, this paper proposes a method for constructing a battery pack health index based on the difference among the discharge intervals of the cells in the pack. First, the discharge interval of each cell is obtained using the online monitorable voltage of each cell in the pack. Then, the Euclidean distance is used to quantify the differences among the discharge intervals of the cells. Finally, the factor analysis method is used to perform fusion to obtain the battery pack health index. There is no parameter selection process for any of the above operations and there is no need to consider the operating conditions. The proposed method has now been tested in this paper using laboratory test data, and the absolute values of the pearson correlation coefficients of the health index obtained are better than 0.93 for each of the conditions tested.