Abstract: A trajectory privacy preserving algorithm based on trajectory direction is proposed. In privacypreserving, dummy is an effective method that was widely used in locationbased service, but in the continuous motion of the user, the user may sends a service request to the position in any point of a track, how to make the dummy looks more realistic is a huge challenge. To solve this problem, we proposed a Trajectory Privacy Preserving Algorithm Based on Trajectory Direction (TPPATD). The middle server can generate different dummies according to different privacy requirements. So as to reduce the probability of the user trajectory being exposed. The experimental results show that, compared with the traditional method of Random Pattern Scheme, Trajectory Privacy Preserving Algorithm Based on Trajectory Direction can generate more dummies when user’s privacy requirements is higher, and dummies are more realistic with the true trajectory.