In this paper, the cell spacing distribution of the battery pack in the parallel air-cooled BTMS is designed to improve the cooling efficiency of the system. The flow resistance network model is used to calculate the airflow rates in the cooling channels. A modification factor is introduced to reduce the error of the model. The effectiveness of the model is verified by the computational fluid dynamics (CFD) calculation, and the CFD method is validated by the experimental air-cooled system with aluminum blocks. Combining with the improved flow resistance network model, an optimization strategy is adopted to optimize the battery cell spacings for the homogenization of the airflow rates among the cooling channels. Then an adjustment coefficient is introduced to adjust the airflow rates in the cooling channels for more uniform cell temperatures. The results of typical numerical cases indicate that the cooling efficiency of the BTMS is improved remarkably after the cell spacing optimization using the developed strategy. Compared to the original BTMS, the maximum temperature of the battery pack for the optimized BTMS is reduced by about 4.0 K, and the maximum cell temperature difference is reduced by more than 69% for various inlet airflow rates. Compared to the optimized BTMS in previous study, the maximum cell temperature difference for the present optimized BTMS is reduced by more than 25% for various inlet airflow rates. (C) 2018 Elsevier Ltd. All rights reserved.