According to the problems that the calculating amount is large, the precision is low and the calculation results are unstable in the distribution state estimation, an improved state estimation method based on artificial bee colony (ABC) algorithm is proposed. The state estimation objective function of distribution network is built firstly depending on network topology and measurement configuration. Secondly, the objective function is optimized by using the ABC algorithm. Thirdly, the fitness function of the current optimal solution is constructed, which can be used as the condition of iteration stopping rule,in order to achieve global optimal solution with high efficiency. The 16-bus test system built by China Electric Power Research Institute is applied to verify the proposed method. The experimental results show that the improved method can estimate the magnitude and the phase of node voltage. Besides, the convergence speed of the proposed method has nearly doubled compared to the conventional artificial bee colony algorithm. The mean absolute error of the proposed method is dropped by 60% and 50% compared with weighted least squares method and the equivalent power transformation method, respectively.