Extracellular vesicles (EVs) carry various informative components, including signaling proteins, transcriptional regulators, lipids, and nucleic acids. These components are utilized for cell-cell communication between donor and recipient cells. EVs have shown great promise as pharmaceutical-targeting vesicles and have attracted the attention of researchers in the fields of biological and medical science because of their importance as diagnostic and prognostic markers. However, the isolation and purification of EVs from cell-cultured media remain challenging. Ultracentrifugation is the most widely used method, but it requires specialized and expensive equipment. In the present study, we proposed a novel methodology to isolate EVs using a simple and convenient method, i.e., an EV catch-and-release isolation system (EV-CaRiS) using a net-charge invertible curvature-sensing peptide (NIC). Curvature-sensing peptides recognize vesicles by binding to lipid-packing defects on highly curved membranes regardless of the expression levels of biomarkers. NIC was newly designed to reversibly capture and release EVs in a pH-dependent manner. NIC allowed us to achieve reproducible EV isolation from three human cell lines on resin using a batch method and single-particle imaging of EVs containing the ubiquitous exosome markers CD63 and CD81 by total internal reflection fluorescence microscopy (TIRFM). EV-CaRiS was demonstrated as a simple and convenient methodology for EV isolation, and NIC is promising for applications in the single-particle analysis of EVs.