The mission of the future parcel delivery will be performed by UAV. However, the GNSS localization in urban area usually experience the notorious multipath effect and non-line-of-sight (NLOS) reception which could potential cause about 50 meters of positioning error. This misleading localization result can be hazardous for UAV applications in the GNSS-challenged areas. However, due to the complexity of multipath, there is no general solution to eliminate the effect. A solution for guiding unmanned aerial vehicle (UAV) operation in the urban area is to plan an optimal route that smartly avoided the area with the dangerous multipath effect. To achieve this goal, the impact of multipath effect in terms of positioning error at different locations must be understood. One method is to simulate the reflection route by ray-tracing algorithm with the aids of predicted satellite positions and the widely available 3D building model. Thus, the multipath effect in pseudorange domain can be simulated using the reflection route and multipath noise envelope according a correlator design. By the reconstructing the multipath-biased pseudorange, the simulated positioning error could be obtained using least square positioning algorithm. Thus, the GNSS error distribution of a wide target area can be further constructed. With the positioning error distribution and 3D building models, an optimal path that avoided not only obstacles but also high multipath effect area can be planned. Both new A* and potential field path planning algorithms are developed to combine with the GNSS error distribution. For the former one, this paper designs a new cost function to consider both the distance to destination and the positioning error at each grid. For the potential field algorithm, a new repulsive field considered both obstacles and high positioning error is developed. By comparing the conventional and the proposed path planning algorithms, the proposed methods can plan paths with less positioning error, namely safer routes for UAV in urban areas.