Purpose: The purpose of this study was to develop an empirically based, psychometrically sound instrument for the assessment of grief in caregivers of persons with Alzheimer's disease. Design and Methods: A total of 184 statements addressing personal grief reactions were obtained from 45 adult child and,42 spouse caregivers in 16 focus groups representing early, middle, late, and post-death stages. These were presented to a second sample of 83 adult child and 83 spouse caregivers, who rated them according to their current experience. Also administered were the Beck Depression Inventory, the Anticipatory Grief Scale, the Caregiver Strain Index, the Caregiver Well-Being Scale-Basic Needs, and the Perceived Social Support-Family Questionnaire. Results: Factor analysis resulted in a 50-item scale containing three factors: Personal Sacrifice Burden, Heartfelt Sadness and Longing, and Worry and Felt Isolation. Cronbach's alpha scores ranged from .90 to .96, indicating high internal consistency reliability for each factor as well as for their combined total. Correlations of each factor with other measures provide evidence of validity, especially when assessed in the context of the caregiver grief model. Implications: Results suggest that caregiver grief is neither a unitary nor a static construct and that a scale suck as this may be appropriate for use in supportive, clinical, and research settings.