The demographic change will lead to an increase in the incidence of falls in the elderly. Technological progress allows for unobtrusive physical activity measurement with miniature sensors, e.g. accelerometers. Yet it is unclear which activities or activity patterns are associated with an increased fall risk. The aim of the research for this paper is to identify daily physical activities associated with a high fall risk. A one-year follow-up study was conducted with n=50 geriatric patients who took part in a telephone interview to assess fall events, their consequences and a set of daily physical activities. Descriptive analysis of the data shows that there are marked differences between fallers (n=21) and non-fallers (n=29) in the overall activity level, the amount of shopping activity and associated locomotion, and in the intensity of light household work. The results confirm that there are differences in typical daily activities between fallers and non-fallers that may be used as parameters to enhance fall prediction models.