Using traditional assessment procedures, prior research demonstrated that deficiencies in gait and balance occur in the later stages of dementia. We tested the hypothesis that an automated system capable of detecting path tortuosity (irregular movement) in elders would show that greater tortuosity was associated with greater cognitive impairment, potentially allowing early detection of dementia over time as tortuosity levels slowly increased. An ultra-wideband sensor network using wireless transponders measured daytime locomotion to an accuracy of 20 cm in 14 elderly residents in an assisted living facility (ALF) as they traversed a shared living area while performing daily activities such as going to a dining area, conversing and watching television. Transponder location was updated at 0.4 sec intervals while in motion and revealed large individual differences in activity patterns. Fractal dimension (Fractal D), a measure of movement path tortuosity (directed vs. irregular or apparently aimless locomotion) was significantly and negatively correlated with cognitive status as measured by the Mini Mental State Examination administered to each participant at the study's end. Previous studies of locomotion in laboratory settings that have demonstrated gait variability increases with poor cognitive status have necessarily controlled various components of gait. The present results demonstrate that directional changes and other locomotion components can be studied by monitoring free movements in normal living settings over time. Implications for assessment and management of dementia-related wandering are discussed.