Abstract Anti-Sm antibodies, identified in 1966 by Tan and Kunkel, are highly specific serological markers for systemic lupus erythrematosus (SLE). Anti-Sm reactivity is found in 5–30% of SLE patients, depending on the autoantibody detection system and the racial background of the SLE population. The Sm autoantigen complex comprises at least nine different polypeptides. All of these core proteins can serve as targets of the anti-Sm B-cell response, but most frequently the B and D polypeptides are involved. Because the BB'Sm proteins share cross-reactive epitopes (PPPGMRPP) with U1 specific ribonucleoproteins, which are more frequently targeted by antibodies that are present in patients with mixed connective tissue disease, the SmD polypeptides are regarded as the Sm autoantigens that are most specific to SLE. It was recently shown that the polypeptides D1, D3 and BB' contain symmetrical dimethylarginine, which is a component of a major autoepitope within the carboxyl-terminus of SmD1. In one of those studies, a synthetic dimethylated peptide of SmD1 (amino acids 95–119) exhibited significantly increased immunoreactivity as compared with unmodified SmD1 peptide. Using immobilized peptides, we confirmed that the dimethylated arginine residues play an essential role in the formation of major SmD1 and SmD3 autoepitopes. Moreover, we demonstrated that one particular peptide of SmD3 represents a more sensitive and more reliable substrate for the detection of a subclass of anti-Sm antibodies. Twenty-eight out of 176 (15.9%) SLE patients but only one out of 449 (0.2%) control individuals tested positive for the anti-SmD3 peptide (SMP) antibodies in a new ELISA system. These data indicate that anti-SMP antibodies are exclusively present in sera from SLE patients. Thus, anti-SMP detection using ELISA represents a new serological marker with which to diagnose and discriminate between systemic autoimmune disorders.