Abstract.Frequency stability of islanded distribution system is a topic of interest, due to the significant penetration of distributed generation (DG). This paper proposes a new under frequency load shedding technique, which address the over-shedding issue caused by fixed priority. The proposed scheme consists of two units; first, the power imbalance estimation unit (PIEU), which use adaptive neuro-fuzzy inference system (ANFIS) to estimates the power imbalance. Second, the load shed unit (LSU), which is based on binary evolutionary programming (BEP) technique to shed the optimal loads. To validate the performance of the proposed UFLS scheme, different simulation studies have been conducted using PSCAD software. Moreover, the response of the proposed scheme is compared with another UFLS scheme having fixed priority loads. From the simulation results, the proposed scheme can shed the optimal loads, which is not achieved in the existing load shedding schemes. IntroductionRecently, controlling the frequency within the permissible limits during islanding operation is the most important technical challenge being investigated worldwide. Hence, alternative techniques are required to enhance the reliability of today's complex distribution network. This can be mitigated by using accurate load shedding (UFLS) scheme, which balances the system. Accurate load shedding depends upon two main factors; accurate estimation of power imbalance and accurate amount of load to be shed. In the literature, several load shedding schemes have been proposed: conventional, adaptive, and intelligent. The conventional UFLS is the most applicable in this case [1][2][3]. In this technique, a pre-determined load values will be shed based on frequency value. Although this technique is very simple, it is suffering from over-shedding issue. For this reason, the adaptive UFLS technique comes with advantage of using swing equation to calculate the imbalance of power. Further efforts to accurately estimate the power imbalance were the application of computational intelligence based techniques such as artificial neural network (ANN) [4][5][6], fuzzy logic control [7-9], particle swarm optimization technique (PSO) [10,11], and genetic algorithm (GA) [12,13]. From reviewing different load shedding techniques, there is inaccurate load shedding issue since every technique is bounded by fixed priorities shedding load. Fixed load shedding priority is linked to the loads that are shed sequentially based on lookup table which consist of critical and non-critical loads. For this reason,