Recent reviews pinpointed the enormous diversity of proteins found in living organisms, especially in higher eukaryotes. Protein diversity is driven through three main processes: first, at deoxyribonucleic acid (DNA) level (i.e. gene polymorphisms), second, at precursor messenger ribonucleic acid (pre-mRNA) or messenger ribonucleic acid (mRNA) level (i.e. alternative splicing, also termed as differential splicing) and, finally, at the protein level (i.e. PTM). Current proteomic technologies allow the identification, characterization and quantitation of up to several thousands of proteins in a single experiment. Nevertheless, the identification and characterization of protein species using these technologies are still hampered. Here, we review the use of the terms “protein species” and “protein isoform.” We evidence that the appropriate selection of the database used for searches can impede or facilitate the identification of protein species. We also describe examples where protein identification search engines systematically fail in the attribution of protein species. We briefly review the characterization of protein species using proteomic technologies including gel-based, gel-free, bottom-up and top-down analysis and discuss their limitations. As an example, we discuss the theoretical characterization of the two human choline kinase species, α-1 and α-2, sharing the same catalytic activity but generated by alternative splicing on CHKA gene.