In this report, we examined whether immunofluorescent anti-nuclear antibody (FANA) test could be used as an initial screening test in identifying serum samples where subsequent testing for antibodies to extractable nuclear antigens (ENA) and to double-stranded DNA (ds-DNA), would be worthwhile. For this purpose, we retrospectively analysed the results of tests for FANA (on Hep-2 cells, positive FANA titers greater than or equal to 1:80), anti-ENA (antibodies to Ro (SSA), La(SSB), U1 nRNP and Sm, by counterimmunoelectrophoresis) and anti-dsDNA (by enzyme-immunoassay) in the sera of 488 consecutive autoimmune patients. It was found that, among the 243 serum samples with negative FANA, only two each (0.8%), had anti-ENA (weak anti-Ro (SSA) and anti-dsDNA (in low levels), respectively. In contrast, anti-ENA and anti-dsDNA were present in 28.2% and 13.5%, respectively, of the 245 FANA-positive sera (p less than 0.00001). In addition, it was found that the probability of positive tests for antibodies to ENA and to dsDNA increases proportionally to the FANA titer (p less than 0.00001). We conclude that searching for anti-ENA and anti-dsDNA in FANA-negative serum samples is generally unjustifiable.