Exome sequencing of in situ tumor samples reveals that mutated genes can predict the prognosis of patients with T‐cell lymphoma (TCL). However, how tumor mutation burden (TMB) derived from circulating tumor DNA (ctDNA) may stratify TCL patients remains unclear.The plasma ctDNA of 79 newly diagnosed TCL patients from the clinical center is used for targeted exome sequencing, and the exome data of 4035 TCL patients from the Catalogue of Somatic Mutations in Cancer (COSMIC) database is obtained for comparison analysis.TCL patients with higher TMB, as evaluated with a panel of 120 genes (panel‐TMB120), are associated with poor prognosis. More importantly, COX regression analysis identifies a subset of 13 genes in panel‐TMB120, including AP3B1 (Adaptor related protein complex 3 subunit beta 1), ATM (Ataxia‐telangiectasia mutated), BCL6 (B cell lymphoma 6), BRAF (B‐Raf proto‐oncogene, serine/threonine kinase), CDKN2B (Cyclin dependent kinase inhibitor 2B), EPCAM (Epithelial cell adhesion molecule), FBXO11 (F‐box protein 11), JAK1 (Janus kinase 1), MDM2 (Murine double minute 2), NF1 (Neurofibromin 1), STAT5B (Signal transducer and activator of transcription 5B), STAT6 (Signal transducer and activator of transcription 6), and TET2 (Tet methylcytosine dioxygenase 2), which are significantly associated with prognosis. Specifically, higher TMB values calculated with these 13 genes (panel‐TMB13) are able to significantly predict unfavorable prognosis for these patients. Together, panel‐TMB13 and the International Prognostic Index (IPI) are used for risk stratification.Panel‐TMB13 is identified, which can predict poor prognosis for TCL patients carrying higher panel‐TMB13 scores and suggest that panel‐TMB13 may be a potential biomarker for supplement risk stratification of TCL patients.