In chronic lymphocytic leukaemia (CLL) the intracellular signalling response that follows B-cell receptor (BCR) ligation varies between cases. This capacity to signal is linked to established prognostic groups: clones possessing relatively unmutated VH genes (UM-CLL) generally retain the ability to signal, whilst somatically hypermutated clones (M-CLL) are mostly unresponsive. As it is known that proteins associated with lipid rafts are important for the initiation of BCR signalling, we hypothesised that differences in the protein composition of rafts could account for the heterogeneity of BCR responses between UM- and M-CLL clones. Lipid rafts were isolated by density gradient centrifugation of CLL-cell lysates. The buoyant fraction thus obtained was enriched for Lyn kinase and ganglioside GM1, whilst lacking CD45, features consistent with the presence of lipid rafts. Lipid rafts were isolated from 12 CLL clones possessing variable degrees of VH somatic hypermutation and of BCR responsiveness. Raft fractions were subjected to SDS-PAGE and raft-associated proteins visualised by staining with Coomassie Blue. This staining revealed multiple bands with a similar distribution between cases. However case-to-case heterogeneity in the intensity of some protein bands was observed. These differences are likely to reflect true variation in protein expression since equal amounts of lysate were used in each experiment (100 μg of protein/lane). 21 of the visualised protein bands were excised, subjected to tryptic digestion and analysed using MALDI-TOF mass spectrometry. 14 proteins were positively identified with a high degree of confidence with the majority of these being cytoskeletal and structural proteins. One protein in particular was variably expressed in the rafts of different clones and was identified as the F1 subunit of ATP synthase; a finding confirmed by immunoblotting of raft fractions. This differential expression of raft-associated ATP synthase in lipid rafts was associated with the extent of VH gene mutation but not with BCR responsiveness. In five of six UM-CLL clones (VH<5%), raft-associated ATP synthase was detectable, whilst only two of six M-CLL clones (VH>5%) expressed raft-associated ATP synthase. Flow cytometric assessment revealed a uniformly weak expression of ATP synthase on the cell surface of all clones tested. Therefore in those cases lacking raft-associated ATP synthase the protein was present on the cell surface, but excluded from lipid rafts. ATP synthase was initially described as a mitochondrial protein, but has subsequently been identified both on the surface of hepatocytes where it controls cellular lipoprotein uptake, and the surface of lymphocytes where it functions as a ligand for natural killer cells. The role of cell surface ATP synthase in CLL cells is unclear but the presence of this protein in the rafts of somatically unmutated clones could be associated with the poor prognosis of patients in this subgroup and therefore warrants further investigation.