We compared the clinical sensitivity and specificity of three different methods for the detection of serum antiribosomal P protein (anti-P) antibodies in systemic lupus erythematosus (SLE). Sera from 60 unselected SLE patients, 100 healthy subjects and 100 patients with other rheumatic inflammatory diseases were screened for anti-P antibodies by immunoblotting (IB) on P proteins from Raji cells and by two ELISA assays, one using the C-terminal 22 aminoacid long synthetic peptide (C-22) of P proteins, the other using a multiple antigen peptide (MAP) carrying four copies of the C-terminal 13 aminoacid long P peptide.Anti-P antibodies were found in 20% lupus sera by IB, 16.7% by MAP ELISA and 11.7% by C-22 ELISA. The specificity for SLE diagnosis of the three tests in healthy subjects and other rheumatic diseases was: 100% by IB, 100% (vs healthy subjects) and 97% (vs rheumatic diseases) by C-22 ELISA, 100% by MAP ELISA. The agreement between methods was good; differences in concordance rates were restricted to weak positivities. We observed a high concordance in the results of IB and ELISA methods for anti-P antibody detection. IB on P proteins extracted from human lymphoid cells is more sensitive than both ELISAs; IB and MAP ELISA perform better than the C-22 ELISA in determining weakly positive sera.