In order to identify the key factors involved in pedestrian search behavior, we analyzed experimental data using statistical designs. Firstly a GPS-GIS-based search mechanism was used and search time, together with geographical characteristics of the town, luminosity and GPS accuracy with between 25.0m and 61.5m in SPS (Standard Positioning Service) were measured. Then the sense of direction of the subjects was derived from SDQ-S (Sense of Direction Questionnaire-Short Form). Statistical methods of regression and path analysis were applied to the data. Results from the analyses identified geographical properties such as the complexity of the road network and the geographical density of landmarks as being most influential in the pedestrian's search process. The accuracy of the GPS system itself was much less important.