The growth in energy demand, the privatization of the power network, and the efficiency of renewable resources have created an opportunity for power system companies and researchers to take advantage of the grid-connected Distributed Generation (DG) systems. The main issue of the DG systems is islanding. Several researchers carried out their research to resolve the problems of the islanding detection. So, Islanding must be detected quickly to maintain the safe operation of the utility grid, to protect DG units, consumer appliances, and line staff, etc. In this paper, a new, and efficient passive islanding technique has been proposed for a grid-connected hybrid distributed generation system. In this technique, the islanding is detected by observing the lower frequency components of the voltage signal measured at the Point of Common Coupling (PCC) using time-domain spectral analysis. The islanding is then identified if the spectral content of the lower frequency components reaches the predetermined threshold value over a given period. The proposed technique detects all types of islanding cases within 0.3 second of the granted delay time even for zero power variation. The effectiveness of the proposed technique has been tested on a wide range of operating conditions and also for different non-islanding situations. It is found that the performance of the proposed technique is quite satisfactory. Furthermore, the response of the technique is fast, independent of system parameters, zero Non-Detection Zone (NDZ), no power quality issues, and its operation are independent of the size and type of DG connected to the utility grid.