Caregivers of persons with dementia (PWD) can experience loss and grief long before the death of the person. Although such experience of caregiver grief is measurable, available scales (such as the Marwit–Meuser Caregiver Grief Inventory, MM-CGI) are lengthy and have overlaps with other caregiving constructs. We developed a briefer scale that captures the essence of caregiver grief—with comparable psychometric properties and total score to MM-CGI, as well as less overlap with other caregiving constructs. Family caregivers of community-dwelling PWD (N = 394) completed questionnaires containing MM-CGI and other caregiving scales. Initially, we split the study samples into two —the derivation sample (n = 179) was used to develop a brief scale that best predicts MM-CGI (using the best-subset approach with tenfold cross-validation), whereas the validation sample (n = 215) verified its actual performance in predicting MM-CGI. Thereafter, we evaluated the derived scale in its reliability and validity, and mapped its scores to MM-CGI using the equipercentile equating method.We derived a 6-item scale, which explained 84.1% of the variability in MM-CGI and had area under the receiver operating characteristic curve of .96 in discriminating high caregiver grief (95% CI: .94–.99). It had single dimension in confirmatory factor analysis (comparative fit index = .98; Tucker–Lewis index = .97) and maintained good psychometric properties similar to those of MM-CGI, while showing lower correlation with caregiver burden and depression. It also had scores that could be mapped to MM-CGI with reasonable precision. We developed the first brief scale with less than 10 items that can conveniently and accurately measure caregiver grief, which opens the way for grief-related interventions in clinical care. Notably, this 6-item scale was developed using rigorous methods and demonstrated consistent evidence of capturing the essence of caregiver grief.