A high-precision and high-efficiency reduced-dimension direction of arrival (DOA) estimation algorithm based on an L-shaped array for the problems of large computation and high cost of achieving two-dimensional (2D) DOA estimation by 2D multiple signal classification (MUSIC) algorithm under various complex arrays. The algorithm makes full use of the structural characteristics of the L-shaped array to decompose the uniform L-shaped array into two uniform linear arrays. These two arrays are respectively searched in one-dimension (1D) to estimate the angles between the source and the x-axis and y-axis, and then the 2D DOA estimation is obtained according to the geometric relationship, which greatly reduces the amount of computation. Furthermore, the algorithm increases the utilization of noise subspace information, which not only realizes the automatic pairing of direction angle and elevation angle, but also improves the estimation accuracy. In order to further reduce the complexity and improve the estimation performance, this paper also puts forward the root finding method instead of 1D search and uses a fast angle matching method to accurately match angles. Simulation results show the feasibility of the proposed algorithm.