Objective To explore the predictive value of carotid ultrasound combined with ABCD3-I score and serum microRNA-146a (miR-146a) for secondary cerebral infarction (CI) after transient ischemic attack (TIA). Methods A total of 90 patients with TIA admitted to the hospital were enrolled between January 2020 and December 2022. According to presence or absence of secondary CI within 30d after TIA, they were divided into CI group (42 cases) and non-CI group (48 cases). The clinical data, severity of carotid stenosis, ABCD3-I score and serum miR-146a at admission in different groups were compared. The correlation among the severity of carotid stenosis, ABCD3-I score and serum miR-146a in CI group was analyzed. The risk factors of secondary CI after TIA were analyzed by Logistic regression analysis. The predictive efficiency of ABCD3-I score, carotid stenosis and serum miR-146a for secondary CI was detected by receiver operating characteristic (ROC) curves. Results The severe carotid stenosis rate, ABCD3-I score and serum miR-146a in CI group were significantly higher than those in non-CI group (P<0.05). Spearman correlation analysis showed that ABCD3-I score and serum miR-146a were positively correlated with the severity of carotid stenosis (P<0.05). Pearson correlation analysis showed that ABCD3-I score was positively correlated with serum miR-146a (P<0.05). Logistic regression analysis showed that age, TIA attacks ≥3 times within 3 months, severe carotid stenosis, ABCD3-I score ≥8.33 points and serum miR-146a ≥1.79 were independent risk factors of secondary CI after TIA (P<0.05). ROC curves analysis showed that the area under the curve (AUC) values of carotid stenosis severity, serum miR-146a and ABCD3-I score for predicting secondary CI were 0.820, 0.889 and 0.854, respectively. Conclusion Carotid ultrasound, ABCD3-I score and serum miR-146a all have good predictive efficiency for secondary CI after TIA, and serum miR-146a has the highest predictive efficiency.