BACKGROUND: The serum albumin-to-globulin ratio (A/G) has been widely used as a biomarker to assess inflammation, immunity, and nutritional status. However, relatively few studies have been conducted on the predictive value of serum A/G in stroke. Therefore, this study aimed to evaluate the correlation between serum A/G levels and stroke prognosis, to provide a new reference for risk assessment and management of stroke patients. METHODS: Data were sourced from the National Health and Nutrition Examination Survey (NHANES) for 2009-2020. The study utilized questionnaire responses and 24-h dietary recall interviews. Participants were stratified by serum albumin/globulin (A/G) ratios into tertiles. Multivariable logistic regression, curve fitting, subgroup analyses, and interaction tests were conducted to assess the associations with serum A/G ratios. RESULTS: Of the 82,298 participants initially considered, 52,119 had complete data and no history of stroke, albumin, or globulin deficiency, which were included in the analysis. We observed a decrease in stroke incidence with increasing A/G ratios. Higher A/G ratios were also associated with lower incidences of moderate exercise, diabetes, and coronary heart disease. The relationship between A/G ratios and stroke was moderated by covariates such as gender, hypertension, diabetes, smoking, and body mass index. CONCLUSION: In the US population, serum A/G ratios positively correlate with stroke incidence. Serum A/G could be a simple and economical marker for identifying stroke risk in the population, though further prospective studies are required to validate these findings.