Recently, several researchers have studied a vector-tracking loop to improve tracking performance for Global Positioning System receivers. The advantage of the vector-tracking loop is that noise is reduced in all of the tracking channels. In addition, the vector-tracking loop can operate successfully when the independent scalar tracking loop approach fails. However, the vector-tracking loop is vulnerable to low-quality signals because all tracking channels are dependent on each other. To solve this problem, this paper proposes an adaptive vector-tracking loop based on a linear local filter and an adaptive navigation filter. The covariance rescaling method is used to lessen the effect of low-quality signals. The proposed adaptive vector-tracking loop shows a better tracking performance compared to the conventional vector-tracking loop that is used when satellite signals are low quality.