Abstract: The direction-finding method using weighted subspace sparse recovery for the underwater acoustic vector sensor array (UAVSA-WSSR) is proposed in this paper, which can be used for the high-resolution azimuth estimation of underwater targets. The proposed method obtains noise-normalized UAVSA data by estimating the noise power of the pressure channels and the particle velocity channels, and uses weighted subspace fitting and reweighted L₁-norm minimization techniques to accurately estimate the azimuths of the targets. The proposed method improves the azimuth resolution performance by noise power weighting and weighted subspace sparse recovery. Simulation results demonstrate that the proposed method performs better than the multiple signal classification (MUSIC) method, the augmented subspace MUSIC (AS-MUSIC) method and the sparse signal reconstruction based on the singular value decomposition (L₁-SVD) in terms of resolution and estimation accuracy. Experimental results of the anechoic pool further validate that the resolution of the proposed method is superior to the other three methods in the presence of one strong target and one weak target.