The goal of many changes in care-environments for people with dementia is to create an environment where the resident can feel at home. Nursing-home residents should be able to continue with dwelling in a normal way. A crucial experience in dwelling is privacy (Altman, 1975). The measurement of privacy is a complicated task. Residents with dementia represent a special user-group for which self-report does not seem to be the best method to find out more about the privacy they experience (Brown, 1992). Current methods to study privacy in the nursing-home mainly focus on the characteristics of the physical environment (Day, et. al., 2000). Research that assesses the privacy people experience is scarce (Munnecom, 2002). Goal of this study was to develop a privacy-measure for nursing-home residents with dementia, based on their behavior. The theory of goal directed behavior plays an important role in the development of the privacy-measure (Kaiser & Wilson, in press). In order to achieve a goal (decrease a lack of privacy) a person makes use of a range of behaviors. These behaviors differ in the effort it requires to engage in them. Residents that are more motivated to decrease their lack of privacy are more likely to engage in behaviors that take more effort. In this study a range of behaviors is selected from privacy-literature (Altman, 1975; Newell, 1995; Pedersen, 1999; Vinsel et. al., 1980). By applying the Rasch model (see Bond& Fox, 2001), it is possible to find out if the behaviors serve the same goal and to find out how these behaviors differ in the effort it requires to engage in them (Kaiser & Wilson, in press). In other words, we derive information about the lack of privacy residents experience from the individual behavior patterns the residents engage in when they strive to reduce their lack of privacy. A total of 59 residents participated in the study. Data on resident behavior was collected in three different nursing-homes through observations and through a questionnaire that was completed by caregivers who were familiar with the participating residents. The three nursing-homes represent three conditions, which show variations in their possibilities for privacy-regulation in the physical environment and in their care-concept. The questionnaire-measure provided reliable data on the lack of privacy of the participating residents. This means a first promising step was made in the development of a privacy-measure for residents with dementia. Possibilities for improvement include making the formulation of the questionnaire for the caregivers more explicit and testing the measure on a larger sample of residents. In the future the measure can be a useful tool to study the effect of adaptations in environments for people with dementia, and this can provide new knowledge about design-decisions and their relation with the privacy of residents with dementia.