An effective method to estimate the state of health (SOH) of lithium ion batteries is illustrated in this work. This method uses an adaptive transformation of charging curves at different stages of life to quantify the extent of capacity fade and derive a time-based parameter to enable an accurate SOH estimation. This approach is easy for practical implementation and universal to chemistry or cell geometry, with minimal demand of learning. With a typical constant current-constant voltage (CC-CV) charging method for a lithium ion battery, this approach uses an equivalent circuit model to characterize the CC portion of the charging curve and derive a transformation function and a time-based parameter to estimate SOH at any stage of life via a nonlinear least squares method to identify model parameters. The SOH estimation errors (discrepancy between estimated and experimental values, denoted as Delta SOH) are under 2% before the end of life in cases shown at 25 degrees C and 60 degrees C and a range of typical discharging rates up to 3C. With different sizes and chemistries, the Delta SOHs are all less than 3%. (C) 2013 Elsevier B.V. All rights reserved.