Extracellular vesicles (EVs) are nanoscale vesicles with a phospholipid bilayer. In the past few decades, EVs have gained more and more attention, which is attributed to their important roles in cell-to-cell communication. They are regarded as promising sources for disease biomarkers and have been explored for applications in early-stage diagnostics, monitoring of disease status, therapeutics and drug delivery. Nevertheless, EVs are a heterogeneous group of vesicles, and include two predominant classes: exosomes and microvesicles. The origins of these vesicles are diverse, which determines their differences in features and functions. To study the diversity of these EV subpopulations, it is essential to elucidate their compositions including proteins, metabolites, etc. Here, we presented a tandem extraction method to obtain metabolites and proteins from the same batch of EVs simultaneously, enabling a multi-omics differential analysis of exosomes and microvesicles in human plasma. As a result, we found 112 different proteins and 50 different metabolites between exosomes and microvesicles, demonstrating the diversity of these EV subpopulations. Furthermore, compared with human plasma, these two major classes of EVs showed distinct metabolome features, which indicated the necessity of analysing the metabolites derived from EVs to obtain a more comprehensive profile of the plasma metabolome, and the potential of EVs as important sources for biomarker screening.