Lupus nephritis (LN) is a major cause of morbidity and mortality in lupus. Renal biopsy is the gold standard for classification of nephritis, but because of its impracticality, new approaches for improving patient prognostication and monitoring treatment efficacy are needed. We aimed to evaluate the potential of metabolic profiling in identifying biomarkers to distinguish disease and monitor treatment efficacy in patients with LN. Serum samples from patients with LN (n = 18) were profiled on NMR-based metabolomics platforms at diagnosis and after 6 months of treatment. LN patients had a different metabolomic fingerprint as compared with healthy controls, with increased lipoproteins and lipids and reduced acetate and amino acids. Using multivariate statistical analysis, we found that the metabolic changes observed in naïve LN patients at diagnosis displayed a variation in the opposite direction upon responding to treatment. Increased levels of lipid metabolites including low- and very-low-density lipoproteins (LDL/VLDL) in LN patients significantly decreased after 6 months of treatment, whereas the serum levels of acetate increased. These levels correlated significantly with SLE Disease Activity Index (SLEDAI 2K), renal SLEDAI, and serum C3 and C4 levels. The result presented in this pilot longitudinal study revealed the reprogramming of metabolome in LN patients on immunosuppressive therapy using NMR-based metabolomics, and thus this approach may be used to monitor the response to treatment.