To further clarify the clinical significance of nDNA antibodies in systemic lupus erythematosus, 70 patients were studied longitudinally over an average period of 3 years each. Serum hemolytic complement (CH50) and nDNA antibodies measured by antigen binding and complement fixation were assessed during episodes of clinically inactive and active disease. Only slightly more than half of episodes of active disease were associated with low CH50 and high nDNA binding, and there was a significant occurrence of high nDNA binding or low CH50 with inactive disease. In contrast, nDNA antibodies measured by complement fixation were never present in the absence of clinical disease. Complement fixing nDNA antibodies were associated mainly with episodes of renal disease, whereas all types of disease activity except those involving predominantly the central nervous system (CNS) showed some correlation with the combined parameters of low CH50 and high nDNA binding. Most episodes of CNS disease occurred without CH50 depression or high nDNA binding. Disease activity was less commonly associated with only low CH50 or high nDNA binding. However, when rash was the predominant manifestation, low CH50 alone was frequently present. The CH50 test correlated with disease activity better than the nDNA antibody binding or complement fixation tests. However, as illustrated by longitudinal studies, a combination of these assays provided the best monitoring of disease activity. Contrary to recent reports, IgG/IgM nDNA antibody ratios did not correlate with disease activity. Studies on isolated nDNA antibody populations provided evidence that differences in complement‐fixing activity were not due to class or subclass composition as previously proposed. Consideration of the above qualifications may result in more effective clinical application of the nDNA antibody test.