Previously, we have shown that anti-DNA can bind to heparan sulphate (HS), a constituent of the glomerular basement membrane (GBM). We hypothesized that binding of anti-DNA to HS in the GBM plays a role in the onset of systemic lupus erythematosus (SLE) nephritis. To test this hypothesis we measured the anti-HS reactivity in cross-sectional and longitudinal studies of SLE patients with or without nephritis. In the transverse serum study single serum samples from 26 SLE patients were studied. We found no correlation between anti-HS reactivity and previously development of nephritis (anti-HS positive: seven out of 16 with history of nephritis, two out of 10 without nephritis). However, six of the seven anti-HS positive sera in the nephritis group were obtained within 1 month of the onset of nephritis, suggesting a temporal relationship between anti-HS reactivity and onset of nephritis. In the longitudinal serum study between six and 16 serum samples were studied from each of 10 SLE-patients. In five out of five episodes of nephritis we found anti-HS reactivity before the onset or exacerbation of the nephritis. In four non-renal manifestations anti-HS reactivity was found in only one episode; in none of the three patients who remained clinically stable did serum samples show anti-HS reactivity. Anti-HS reactivity was only found in sera positive for anti-DNA by Farr assay but the anti-HS titre was not a mere reflection of the reactivity measured in the Farr assay. This indicates that only a subpopulation of anti-DNA can bind to HS. We found a high correlation (r = 0.99) between anti-HS reactivities in plasma and serum and we conclude that anti-HS reactivity in serum samples from SLE patients is not due to in vitro complex formation during clotting. Although further prospective analysis is necessary, our data suggest that measurement of anti-HS reactivity in SLE patients might identify patients at risk for the development of nephritis.