The State of Charge (SOC) of lithium batteries can be estimated effectively by particle Filter (PF). However, it is difficult to accurately obtain the value of the parameter discharge current of the lithium battery in the actual process, which affects the accuracy of the PF for SOC estimation. In order to solve the problem, this paper proposes a particle filter based robust state and parameter estimation (PF-RSPE) method for estimating SOC and discharge current of lithium batteries. The method estimates SOC and the unknown parameter discharge current simultaneously. The unknown parameter is expanded to the states of the model. The kernel smoothing method with the fitness factor β is applied in the procedures of the simultaneous estimation, which improves the performance of the estimation. Finally, the proposed method is applied to the equivalent model of Thevenin battery for simulation verification. The results show that when the parameter discharge current is unknown, the PF-RSPE method can effectively estimate both SOC and discharge current.