We prospectively analyzed the diagnostic sensitivity and specificity as well as the clinical relevance of antinucleosome antibodies in SLE. One hundred and one consecutive SLE patients were followed for 3 years. Three serial serum samples from each patient were tested for antinucleosome antibodies by ELISA (optimum cut-off value 10 U/ml), and for anti-dsDNA antibody (by ELISA and IIF on Crithidia luciliae), and anti-dsDNA avidity (by Scatchard plot analysis). Sera from 100 healthy individuals (HI), 35 patients with systemic sclerosis (SSc), 30 with primary Sjögren's syndrome (SS), 20 with rheumatoid arthritis (RA) and 48 with infectious diseases (ID), were assayed as controls. SLE activity and damage were evaluated using the ECLAM score and the SLICC/ACR index. At baseline, antinucleosome antibodies were found in 87 patients with SLE (86.1%), in 8 patients with SSc (22.8%), in 2 HI (2%), and in 1 ID (2.1%). The sensitivity and specificity of antinucleosome testing for SLE were 86.1% and 95.3%, respectively. The prevalence of antinucleosome antibodies in SLE was significantly higher than that of anti-dsDNA antibodies, with a correlation between them. No relevant relationship was found between antinucleosome antibodies and disease features, including renal involvement, disease activity, and disease damage. During follow-up, no significant variation was observed in antinucleosome level, nor in anti-dsDNA antibody level or avidity. We conclude that antinucleosome antibodies are commonly found in SLE. Low antibody levels can be detected in SSc, whereas medium/high levels are highly specific for SLE. Their clinical relevance during the disease course and utility for monitoring the individual patient seem to be poor.