Several ‘great walls’ are clearly visible in the Durham/UKST Galaxy Redshift Survey (DURS). We make a statistical study of this superlarge-scale structure (SLSS) by applying our core sampling, cluster, inertia tensor and minimal-spanning-tree analyses to the DURS. The results in the main support similar results from the complementary Las Campañas Redshift Survey (LCRS); the DURS is a fully three-dimensional, though shallower, survey, whilst the LCRS was carried out in six thin wedges of space. Because of the one-in-three sparse sampling used for DURS, the galaxy filaments of large-scale structure (LSS) are less clear here; the mean separation of ∼25 h−1 Mpc for the richer filaments is consistent with the LCRS result, but the poorer filaments are not seen in the DURS. In contrast, the analysis clearly picks out SLSS and we find, as with the LCRS, that ∼50 per cent of the galaxies lie within the SLSS in regions with overdensities of 5-10 times the mean galaxy density. It quantitatively demonstrates that SLSS is a major component of large-scale structure in the Universe. The SLSS is also confirmed as having statistical parameters similar to those for a sheet-like object, albeit a very irregular one with a highly inhomogeneous inner structure. The ‘mean-free path’, or average separation between SLSS structures, is found to be Ds≈50 h−1 Mpc. The inertia tensor analysis gives mean lengths, widths and thicknesses of ∼20-40, 10 and 5 h−1 Mpc, respectively, for the clusters of SLSS. In particular, the largest great wall in the DURS is found to have a length of ∼75 h−1 Mpc. Unlike the LCRS, the cluster mass function for the three-dimensional DURS has a high mass ‘tail’; such a ‘tail’ would constitute a quantitative signature for the presence of great walls. Finally, theoretical considerations would suggest that the results support arguments for the large-scale biasing of galaxies with respect to dark matter.