Falls are a major clinical problem in geriatric medicine. Both falls and fear of falling significantly reduce activity and independence of seniors. Systems permitting a person to call help by a button push reduced some problems from falls. This hardly suffices, since with syncope, injury and mental confusion a person may not be able to respond. This project develops a personal alarm device innovatively combining activity, body-positions, and firmware that after a significant fall saves data and automatically calls for help. Despite the clinical significance of falls there has been little effort to develop technology to better describe the phenomenon as it occurs. The proposed ambulatory device also meets this research need by recording activity and body position characteristics associated with a fall. The phase I prototype included an algorithm for detecting falls. It records activity in 3-D plus 6 body positions before, during and following the event. The prototype demonstrated feasibility of concept. Phase II extends the technology to: 10 body positions, improve falls detection, detect behavioral changes associated with falls-risk, waterproof and wirelessly automatically alert the caretaker after a fall. The units will be evaluated on high falls-risk neurological patients. Both research and commercial units will be developed.