Characterisation of blood serum peptides can provide valuable information on physiological and pathological processes. However, the analysis of raw serum samples by MS results in the identification of a limited number of peptides. In order to improve sensitivity, many peptide enrichment methods have been proposed during the last ten years. Here, we present a comparison of fractionation methods aimed to simplify analysis of small proteins and peptides in blood serum, one of the most promising sources of putative biomarkers. Specifically, three methods based on ultrafiltration, differential precipitation, and peptide ligand libraries (ProteoMiner) were evaluated for the enrichment of peptides and low molecular weight proteins, as demonstrated by Tricine SDS-PAGE and subsequent LC-MS/MS (GeLC-MS/MS). As a result, differential solubilisation (DS) allowed the identification of the highest number of peptides. Moreover, the DS method enabled also the quantitative comparison of samples, producing fundamental information in biomarker discovery approaches.