The frequency of anti-U1-RNP and anti-Sm antibodies in systemic lupus erythematosus (SLE) varies according to the technique used for their detection. OBJECTIVE To study the clinical manifestations associated with anti-U1-RNP and anti-Sm antibodies identified by different techniques. METHODS 112 patients diagnosed with SLE were studied. Anti-U1-RNP and anti-Sm antibodies were detected by ELISA with a mixture of recombinant antigens, antigens obtained by immunoaffinity and by immunoblotting and counterimmunoelectrophoresis (CIE) with a saline extract of rabbit thymus. RESULTS Anti-U1-RNP antibodies were detected in 70 (62.5%), 37 (33%) and 16 (14.2%) patients, and anti-Sm antibodies in 43 (38.3%), 36 (32.1%) and 10 (8.9%) patients by ELISA, immunoblotting, and CIE, respectively. Anti-U1-RNP antibodies are associated with edema in the dorsal hand region, sclerodactylia, Raynaud phenomenon and pericarditis (CIE); arthritis, hand edema, Raynaud phenomenon and sclerodactylia (anti-70 KD-U1-RNP and anti-A-U1-RNP by immunoblotting); and a lower incidence of discoid lupus and chronic renal insufficiency by lupus renal disease (ELISA). Anti-Sm antibodies are associated with Raynaud phenomenon and renal disease (CIE); cutaneous vasculitis, Raynaud phenomenon and thrombosis (anti-BB'-Sm immunoblotting); and arthritis and lower incidence of chronic renal insufficiency (ELISA). CONCLUSIONS Anti-U1-RNP and anti-Sm antibodies detected by ELISA do not have the same clinical significance than those detected by immunoblotting or CIE. The clinical manifestations associated with such antibodies in the same SLE population vary according to the detection technique used.