Existing methods for predicting the capacity fade of lithium ion batteries (LIBs) are typically limited to the description of cells under fixed charge/discharge (C/D) cycling conditions. The prediction requires a large amount of experimental data for various cycling environments to construct empirical equations or to obtain parameters in physical-basedmodels. Based on the porous electrode theory, a semi-empirical model with three constant parameters is developed here. It is found that fading data at various temperatures collapse into a single curve on a newly proposed plot. This finding enables us to use a much smaller amount of cycle-life data to determine the values of these parameters, which sufficiently characterize the fading response under a broad range of cycling conditions. The prediction yielded from the model is confirmed by several scenarios with varying cycling conditions, such as temperature, depth of discharge, and C/D rate. A further prediction shows that different average C/D rates, which give rise to different cell-temperature histories, have a strong influence on the capacity degradation of an LIB. By introducing the average C/D rate, we analyze the capacity loss after 1500 cycles with the use of two types of charging protocols, from which two operational treatments are suggested. (C) 2017 The Electrochemical Society. All rights reserved.