Objective To investigate the relationship between the mean corpuscular volume/red blood cell distribution width ratio (MRR) and stroke, providing a new perspective for the risk assessment and early prevention of stroke. Methods The study was based on the complex sampling design of data from the National Health and Nutrition Examination Survey from 1999 to 2018 and included 17434 eligible participants aged 18 years and older. MRR was calculated based on the mean corpuscular volume and red blood cell distribution width, and stroke patients were determined based on the health status questionnaire. According to the quartiles of MRR, the included participants were divided into Q1, Q2, Q3, and Q4 groups. Multivariate logistic regression analysis was used to explore the association between MRR and the incidence of stroke. Official sampling weights, stratification variables, and primary sampling units were incorporated throughout this study for weighting adjustment, and weighted statistical methods suitable for complex samples were adopted for data analysis and intergroup comparisons. Results Among all the participants, there were 625 stroke patients, with a weighted stroke prevalence of 2.7%. The MRR of the overall population were (6.58±0.78) fL/%. The results of the multivariate logistic regression analysis showed that, after adjusting for age, gender, race, education level, diabetes, hyperlipidemia, hypertension, smoking history and drinking history, a high level of MRR was independently associated with a reduced risk of stroke [Q2 vs. Q1: odds ratio (OR)=0.986, 95% confidence interval (CI) (0.976, 0.996), P=0.005; Q3 vs. Q1: OR=0.986, 95%CI (0.975, 0.996), P=0.008; Q4 vs. Q1: OR=0.984, 95%CI (0.974, 0.994), P=0.002]. Conclusions The higher the MRR, the lower the risk of stroke. A higher MRR level provides a protective effect against stroke, and the MRR index has potential value in the prevention and management of stroke.