Abstract: To investigate the relationship between peripheral microRNA (miR) expression of thoracic cancer patients and radiation pneumonitis (RP) during radiotherapy, a total of 18 thoracic cancer patients received radiotherapy in The Second Affiliated Hospital of Soochow University from April 2019 to May 2020 were enrolled. 4 patients were excluded due to the failure of obtaining samples, and 1 patient was excluded for not completing radiotherapy. Therefore, the total of 13 patients was included in this study. Serum samples were collected before and after radiotherapy with 20-30 Gy and 40-50 Gy. The total RNA was extracted and the expression of miR was detected by Agilent Human miR microarray. Then, differentially expressed miRs were identified via T-test and correlation analysis. Next, target miRs of these differentially expressed miRs were predicted by searching miRDB and miRWalk databases, and conducting gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG) analysis. 14 differentially expressed miRs were detected after 20-30 Gy radiotherapy, with 2 up-regulated and 12 down-regulated. Among them, 3 miRs were differentially expressed in a dose-dependent manner after 20-30 Gy and 40-50 Gy radiotherapy. On the other hand, for these 14 differentially expressed miRs, 11 of them change more significantly in patients with ≥2 degree radiation pneumonitis than those without. Further, hsa-miR-19b-3p and hsa-miR-933 differentially express in a dosedependent manner after radiotherapy, and changes more significantly in patients with ≥2 degree radiation pneumonitis than those without. Therefore, hsa-miR-19b-3p and hsa-miR-933 can be identified as potential biomarkers. Finally, target genes of these differentially expressed miRs are predicted to be primarily enriched in the palmitoyltransterase activity, ubiquitin ligase complex, glutamate, Hedgehog, cGMP-PKG and AGE-RAGE signal pathway. The expression of peripheral miR of thoracic cancer patients changes significantly after radiotherapy, and hsa-miR-19b-3p and hsa-miR-933 are screened as potential biomarkers for the prediction and evaluation of RP.