Objective To evaluate the diagnostic value of anti-nucleosome antibody (AnuA) in patients with systemic lupus erythematosus (SLE). Methods We searched MEDLINE, EMbase and The Cochrane Library to identify studies on the diagnostic value of AnuA in patients with SLE. The searching time was from 1990 to 2005. The quality of included studies was evaluated and the data were extracted. The Cochrane Collaboration’s RevMan software was used to analyze heterogeneity, and MetaDisc was used to perform meta-analyses and draw summary receiver operator characteristic curve (SROC). Results Twenty-five studies involving 7 289 patients (2 459 SLE and 5 030 non-SLE patients were diagnosed by gold standard) were included, most of which were poor in quality. The heterogeneity among studies was high (Plt;0.000 01, I2=87.2%). The pooled sensitivity was 64.9%, 95% CI 63.0 to 66.85, and the pooled specificity was 92.6%, 95% CI 91.8% to 93.3%. Area under the SROC was 0.918, SE0.0212. These data suggested that AnuA had a relatively high false negative rate (35.1%) and a relatively low false positive rate (7.4%). Conclusions AnuA has some value in diagnosing SLE and could possibly be used as one of the diagnostic tests for SLE.