An aging society is a reality in developed countries. An aging population requires more health-care workers and facilities. To reduce this social problem, it is worthwhile to develop a wearable computer for elders or patients to watch over them. In this study, we developed a wearable computer, in which accelerometers were installed to detect variations of posture, falls, and gait disability. The advantages of this system include a designated database server in each patient's home, scalability and flexibility to adapt to patient's needs, and full patient access to their own information. As a first step, we adopted this system for healthy young volunteers with or without impediments to validate the system. The results show that this system can successfully detect variations in posture and falls. We also succeeded in real-time automatic gait analysis by using the Hampering Index. The present study gives useful knowledge for the development of a wearable computer to support the care of elders or other patients.