Dementia caregiving is associated with depression, stress, and sleep disturbance. A daily use caregiver sleep survey (DUCSS) was developed to evaluate caregiver sleep. The tool was distributed to 24 informal caregivers and validated using the Rasch model, which indicated that the 17-item survey produced sleep quality measures of sufficient reliability for both group-level and individual-level comparisons (reliability = .87). The sample size was sufficient to provide precise measures of the item’s position along the scale (item difficulty) (reliability = .85), so outcomes associated with sleep quality levels could be evaluated. We observed that the structure of the instrument is unidimensional, meaning the wording does not contain systematic biases peripheral to sleep quality. DUCSS is a useful tool for caregiver assessment and monitoring.