Abstract: The high volatility and uncertainty of distributed generation (DG), such as wind and solar power, lead to increased losses and peak-to-valley differences in distribution networks. The key role of distributed energy storage (DES), predominantly based on electrochemical technologies, in ensuring the safe and economic operation of distribution networks is becoming more prominent. To this end, this study proposes a two-stage adaptive robust optimal placement model for DES in unbalanced active distribution networks, considering DG uncertainty. First, a comprehensive network loss sensitivity metric adapted to multiple scenarios is proposed to enable rational siting of DES in unbalanced active distribution networks. Based on this, a two-stage adaptive robust optimal placement model, including DES planning stage and DES operation stage, is proposed to achieve optimal capacity considering DG uncertainty. The planning stage aims to minimize the investment and maintenance cost while maximize DES scheduling revenue by optimizing DES capacity. The operation stage aims to optimize the DES charging/discharging power and its control strategy to reduce network loss and battery degradation while increasing the peak shaving and valley filling revenue, thereby supporting the realization of the DES charging and discharging revenue objectives in the planning stage. The above two-stage adaptive robust optimization problem is solved using the column and constraint generation algorithm. Finally, the feasibility and superiority of the proposed two-stage adaptive robust optimization configuration method for DES in unbalanced distribution networks are verified through a case study on a real distribution network.