AIMS: Accurate classification of haematological malignancies is a prerequisite for their correct diagnosis, prognosis and therapy. Clear classification of lymphomas is often hindered by the limited number of available cell surface protein markers that are suitable for immunophenotyping. A systematic and quantitative analysis of cell surface proteins is thus required to identify new protein markers on lymphoma subtypes in an unbiased and discovery-driven approach. METHODS: Nine Hodgkin and non-Hodgkin B cell lines of diffuse large cell type and mediastinal type were investigated by cell surface capture (CSC) technology, a mass spectrometry-based method to identify cell surface glycoproteins. Selected proteins are verified by antibody-based methods, including flow cytometry and immunohistochemistry on cell line arrays. RESULTS: A total of 747 predicted transmembrane proteins were identified from all cell lines, including 142 CD (cluster of differentiation) annotated proteins. A group of differentially expressed cell surface glycoproteins between Hodgkin and non-Hodgkin B cell lines was revealed via quantitative CSC technology. In addition to classical and expected CD molecules such as CD20 and CD30, less frequently expressed molecules such as CD2 on Hodgkin lymphoma (HL) cell lines were identified by CSC and verified by immunohistochemistry in cell lines and primary lymphoma tissue. A panel of CSC-identified differentiation glycoprotein candidates is currently under investigation on tissue microarrays (TMAs) from patient samples.