Allostatic load (AL) is a composite measure of cumulative biological damage derived from physiological markers across multiple systems. While previous reports have shown that AL predicts cardiovascular events at early old age, the prognostic value of AL on stroke mortality among the oldest old remains largely unexplored. The objective of this study was to evaluate the association between AL and stroke mortality risk in a birth cohort of older Danish adults from the 1914 Glostrup Aging Study. In total, 330 Danish participants completed a structured questionnaire at baseline (age 80) and were assessed for 10 physiological markers that spanned the cardiovascular (systolic and diastolic blood pressure, heart rate), metabolic (HDL, ratio of total cholesterol/HDL, BMI, triglycerides, serum albumin, blood glucose), and inflammatory (IL6, TNF-alpha) systems. Composite summary scores of AL ranging from 0-10 were computed for each participant using a count-based approach, with high and low cut-points of AL defined at the median. Using the Danish National Civil Death Registry, we ascertained deaths from cardiovascular disease and stroke for all participants and examined survival probability over a 15-year period with Cox regression. Overall, there were 106 cardiovascular disease-related deaths, of which 25 were due to stroke. Older adults with high AL had a 2-fold higher hazard risk of death from stroke than those with low AL (hazard ratio [HR]=1.96, 95% confidence interval [CI]: 0.89 - 4.35), after adjustment for conventional risk factors of stroke. When stratified by sex, men with high AL had about a 6-fold higher risk of death from stroke than those with low AL (adjusted HR= 5.67, 95% CI= 1.66 - 19.4). No association between AL and stroke mortality risk was observed in women (adjusted HR=0.78, 95% CI: 0.23 - 2.63) among whom there were only 6 stroke deaths. In a birth cohort of older Danish adults, AL showed promise as a prognostic tool for stroke mortality risk. Larger studies in more diverse populations are necessary to confirm our findings.