It is very important to estimate the state of charge accurately and to achieve the on-line updating of the battery parameter for the battery management system of electric vehicles. This research aims to develop a novel parameter adaptive method with dead zone for estimation of state of charge and other parameters of lithium batteries. The dead zone refers to the definition of an interval based on the error of the battery model terminal voltage and the measured terminal voltage. When the error is no longer in the interval, the battery parameters are not estimated. Otherwise, the battery parameters are estimated. This research can be summarized as follows. First, the proposed method is applied to estimate all battery parameters including battery capacity, battery impedance and open circuit voltage. Second, the use of dead zone solves the problem of poor robustness of parameter adaptive algorithm when the initial state error is large, and the problem of instability. Finally, the experimental results indicate that the proposed method can achieve estimation accuracy with an error of 1%. Moreover, compared with the method used the same sampling time for estimating battery state and parameters, the dead zone method reduces the computation.