Abstract: A simple method based on a single 3-dimensional vector sensor is proposed to localize and track a surface target.The maximum likelihood estimation of the azimuth and elevation angles is obtained by the averaged intensity estimator.When the vector sensor depth is known, the target position and velocity can be estimated from the target direction sequence.However, as the signal-noise ratio(SNR) is reduced, the accuracy of the azimuth and elevation angles is deteriorated, which results in the fallibility of the target position and velocity estimations.For linearly moving targets, a linear least square method is used to fit the target track, and thus the velocity estimation is improved.The simulation results show that the curve-fit method enhances the accuracy of both the localization and velocity estimation results.The proposed method is simply and low-cost in computation consumption, suitable for passive sonar buoy and torpedo systems, and increases the reliability of the surface target motion analysis based on a single vector sensor.