This report develops a novel lectin-based affinity chromatography (LAC) method for extracellular vesicles (EVs) and demonstrates the possibility of subclassification based on their surface glycan.　EVs are lipid bilayer vesicles enclosing various biomolecules and represent their promising potential as a sensitive biomarker for detecting and monitoring various diseases. However, EVs have heterogeneous molecular compositions, even from identical donor cells by the same purification methods, which is a significant obstacle to elucidating objective biological functions. Therefore, subclassification based on their molecular compositions has been an urgent subject for addressing their heterogeneity. Here, as a novel LAC platform for EVs, we utilized a spongy-like monolithic polymer (named spongy monolith, SPM), which consists of poly(ethylene-co-glycidyl methacrylate) with continuous micropores that allowed efficient in-situ protein reaction with the epoxy groups. Two distinct lectins with different specificity, Sambucus sieboldiana agglutinin and concanavalin A, were effectively immobilized on SPM with remaining binding activity. Moreover, the large flow-through pores (>10 μm) of the SPM allowed high recovery rates of liposomal nanoparticles as a model of EVs. Finally, we employed lectin-immobilized SPMs for the subclassification of EVs based on their surface glycan structures and demonstrated their different subpopulations by proteome profiling.