In an unselected population of 1390 consecutive Albanian patients with rheumatic diseases (RD) and other miscellaneous non-rheumatic diseases (MNRD), for whom antinuclear antibody (ANA) testing was requested, we calculated the diagnostic sensitivity, specificity and positive predictive value (PPV) of ANA positive results, ANA titres over 1:100, anti-native DNA (nDNA), anti-Sm, anti-U1 RNP, anti-SSA (Ro) anti-SSB (La) and anti-non-identified extractable nuclear antigen (NIENA) antibodies. The PPVs of these ANA types were found to be appreciable only for systemic lupus erythematosus (SLE); only the positive predictive value of ANA for SLE (26.4%) was lower than that for RA (34.3%). The anti-snRNP (Sm/U1RNP) positive SLE patients were more likely to have over 4 of the ARA criteria for SLE, ANA titres over 1:100, and anti-nDNA antibodies, in contrast with the anti-snRNP negative subgroup. On the other hand, the anti-ENA positive and anti-nDNA positive SLE patients generally showed higher frequencies of renal disease, over 4 of the criteria for SLE and ANA titres over 1:100, compared to anti-ENA positive and anti-nDNA negative patients. Our data suggest that the association of anti-snRNP antibodies with a more severe form of SLE is not to be attributed to these antibodies themselves, but rather to their close association with the concomitant presence of anti-nDNA antibodies.