The treatment of systemic lupus erythematosus (SLE) nephritis, while effective, is associated with significant morbidity and mortality. These side effects could be mitigated if the onset, severity, and response of renal flare could be predicted, and therapy modified accordingly. In this review, an approach to derive prediction equations of SLE nephritis flare is discussed. Integral to generating such prediction equations is the identification of biomarkers of lupus nephritis that can serve as input variables for modeling flare. The use of urine as a source of SLE nephritis biomarkers is described, and the results of urine biomarker discovery studies using candidate and proteomic approaches are presented.