We have developed a new technology to predict capacity fading of lithium-ion cells using the analysis of their discharge curves. A cell capacity is deduced from the cell voltage window and the cell discharge curve, which was obtained by superimposing discharge curves of the positive and negative electrodes. This new technology consists of three steps: first, to analyze cell discharge curves measured during the calendar life test in order to evaluate deterioration indicators such as the positive and negative usable capacities and the irreversible capacities caused by the side reactions: secondly, to formulate these indicators as the functions of the test time; lastly, to produce the cell discharge curves by using these functions and deduce corresponding cell capacity, over storage days for prediction. Prediction of the faded capacities of the cells after a 960 day calendar life test was attempted with the data up to 360 days using the technology described above and the conventional method by extrapolating capacity-fading data in accordance with square-root time rule. Consequently, the proposed technology predicted the faded capacities more precisely than the conventional method, particularly under severe conditions. (C) 2011 Elsevier B.V. All rights reserved.