Objective To investigate the association between right-to-left shunt (RLS) and the severity of white matter lesions (WML). Methods Patients hospitalized in the Department of Neurology of Northern Jiangsu People’s Hospital from January 2022 to July 2025 who underwent contrast-enhanced transcranial doppler were consecutively enrolled. Baseline information and clinical characteristics were collected. All enrolled patients underwent contrast-enhanced transcranial doppler and brain magnetic resonance imaging. WML was defined as hyperintense signals on T2-weighted and fluid-attenuated inversion recovery (FLAIR) images in the periventricular or deep white matter regions. WML severity was assessed using the Fazekas scale. Mild WML was defined as 0-3 points, and severe WML as 4-6 points. RLS was defined as the presence of at least one microembolic signal in the unilateral middle cerebral artery monitored under the single-channel double-depth mode during contrast-enhanced transcranial doppler. Results A total of 125 hospitalized patients who completed contrast-enhanced transcranial doppler were included, including 93 patients in the mild WML group (74.4%) and 32 patients in the severe WML group (25.6%). RLS was positive in 32 patients (25.6%). Univariate analysis using the Mann-Whitney U test showed that, compared with the mild WML group, the severe WML group had a higher proportion of RLS-positive patients (P<0.01), significantly higher neutrophil count (P=0.003), and significantly higher neutrophil-to-lymphocyte ratio (NLR)(P=0.004). The Mini-Mental State Examination score was significantly lower in the severe WML group (P<0.01). Binary logistic regression with RLS positivity, NLR, age, hypertension, and diabetes as independent variables showed that RLS positivity (OR=39.290, 95%CI: 9.54-161.74, P<0.01), age (OR=1.090, 95%CI: 1.02-1.16, P=0.011), and hypertension (OR=3.749, 95%CI: 1.02-13.84, P=0.047) were independent risk factors for WML severity. Single-factor ROC curve analysis showed that the area under the curve (AUC) of RLS positivity for predicting WML severity was 0.76 (95%CI: 0.663-0.875, P<0.01), indicating good predictive performance, whereas age (AUC=0.557) and hypertension (AUC=0.528) showed relatively poor independent predictive performance. A combined predictive model was constructed by incorporating RLS positivity, age, and hypertension into binary logistic regression, and an ROC curve was generated based on the predicted probability. The results showed an AUC of 0.831(95%CI: 0.743-0.920, P<0.01), with an optimal cutoff value of 0.431, sensitivity of 0.656, and specificity of 0.946, suggesting that the multi-indicator combined model has good predictive value for WML severity.Conclusion The severity of WML may be associated with the presence of RLS.