Human blood contains proteins secreted by various organs, but there is no consensus on whether serum or plasma is preferable for proteome studies. Data-independent acquisition (DIA) mass spectrometry (MS)-based proteomics has proven to be a powerful tool, capable of quantifying thousands of proteins in a single LC-MS/MS experiment, facilitating identification of potential marker and elucidation of biological processes. Here, we profiled the proteome data of ten paired plasma and serum samples in the initial sample set. Functional analysis revealed similarities and differences in biological functions, and the preference for different organs between serum and plasma. Furthermore, comparative proteomic analysis highlighted the different proteomic characteristics, with plasma-overrepresented pathways related to the phagosome and immune while serum-overrepresented pathways were associated with amino acid metabolism, which were further validated by the follow-up sample set composed of eight paired plasma and serum samples. We have detected potential markers in plasma and serum for various cancers and explored their association with prognosis using data from the TCGA pan-cancer cohort and HPA database. Further assessment is required to validate the reproducibility of quantification for these markers. Overall, this study highlights the commonality and specificity of plasma and serum at the molecular level, underscoring their respective utilities in biological exploration and clinical applications.