Esophageal cancer is a malignant tumor with two-thirds of patients having a local recurrence or distant metastasis. To date, diagnostic biomarkers with high sensitivity and specificity are lacking. Extracellular vesicles (EVs) have shown their potential values as disease biomarkers as they carry specific proteins and RNAs derived from cancer cells. In this study, we investigate ESCC precision diagnostics from the insights of circulating EVs, and integrate the ultrafast EV isolation approach (EXODUS) and ELISA for fast detection and screening of ESCC patients. First, we isolate and characterize the high-purity plasma EVs with EXODUS and identify 401 proteins and 372 proteins from ESCC patient and healthy individuals, respectively. Further looking into the differentially expressed proteins (DEPs) of ESCC patients and enriched KEGG pathways, we discover EV-CD14 as a potential diagnostic biomarker for ESCC, which has been further validated as a significantly differentially expressed protein by Western Blot and immunogold labelling TEM. For fast screening and detection of ESCC towards clinical applications, we apply ELISA method to diagnose ESCC from 60 clinical samples based on circulating EV-CD14, which shows a high AUC value up to 96.0% for detection of ESCC in a test set (30 samples), and displays a high accuracy rate up to 90% for prediction of ESCC in a screening test (30 samples). Our results suggest that the circulating EV-CD14 may highly be related to the initiation and progression of ESCC, providing a novel method for the diagnosis and prognosis of ESCC towards clinical translations.Copyright © 2023 Elsevier B.V. All rights reserved.