We have developed a 3-D pedestrian guide system using a dead-reckoning method based on walking locomotion measured by wearable self-contained sensors in combination with the error compensation method with GPS, RFID, and map information. This system is promising for making a ubiquitous information society more sustainable since it can provide indoor and outdoor location-based information services not only seamlessly, but pervasively without excessive dependence on infrastructure-side sensors. A pilot user study of the system was conducted at the Science Museum (five floors, 2500 to 2700 m²/floor) in cooperation with twenty-two subjects. This was an unprecedented indoor 3-D guide experiment in terms of the scale. In addition, we have developed an augmented-reality log browser capable of retrieving and visualizing the trajectory, voice, and video of each subject in a cross-sectional way using queries such as time, place, subject ID, etc.