Patients with lupus erythematosus (LE) that have discoid lesions who fulfill the four diagnostic criteria of systemic lupus erythematosus (SLE) with only mucocutaneous findings and antinuclear antibody (ANA) positivity were classified as borderline SLE in the literature. Objective of this study was to determine the place of borderline SLE with discoid lesions on the LE spectrum according to the lupus band test (LBT). Lesional and sun-protected non-lesional (SPNL) skin LBTs of 94 patients with LE that had discoid lesions were retrospectively evaluated. Firstly, patients were divided into two main groups: discoid LE (DLE; group A) and SLE (Group B); three subgroups were then classified as DLE (Group A), borderline SLE (Group B1) and SLE (Group B2) using another method. Each group had its own comparisons. Immunoreactant (IR) deposition was observed on the lesional skin in all patients and on the SPNL skin in 42 (44.7%). In patients with borderline SLE, the deposition of IgM was lower on the lesional LBTs, whereas isolated IgG was higher than SLE; thus, it shows similarity with DLE. Additionally, it was also closer to DLE because of the low deposition of C3, multiple IRs, and a double conjugate of IRs on the SPNL skin. However, it showed similarity with SLE in the high percentage of LBT positivity and more immunoglobulin M (IgM) and immunoglobulin G (IgG) deposition on the SPNL skin. The deposition of multiple conjugates on SPNL skin in patients with LE with discoid lesions may reflect systemic involvement. Despite the fact that LBT positivity on SPNL skin in borderline SLE was higher than DLE, less deposition of multiple conjugates compared to SLE indicates that the classification of borderline SLE with discoid lesions in the LE spectrum is questionable.