Objective. To determine if measurement of serum complement split products (C4d, Bb, C5b-9) is better than conventional C3 and C4 measurements in distinguishing patients with varying degrees of lupus disease activity, and to determine if the presence of C3d in urine is helpful in distinguishing lupus patients with from those without early lupus nephritis.Methods. Lupus disease activity was prospectively determined at 3 consecutive visits an average of 4 months apart, using the Systemic Lupus Activity Measure (SLAM), the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI), and physician global assessment (PGA). Blood samples were evaluated for the presence of C4d, Bb, and C5b-9 by quantitative microsassay plate enzyme immunoassay at each patient visit. We characterized urinary excretion of C3 fragments (with attention to C3d) by sodium dodecyl sulfate-polyacrylamide gel electrophoresis with Western blotting.Results. Thirty-one SLE patients were enrolled in the study. The mean SLAM score and the mean SLEDAI score each correlated well with the PGA at all 3 visits. A SLAM score of 6 and a SLEDAI score of 4 had the best overall sensitivity and specificity for predicting moderate-to-severe disease activity by PGA (100% and 73%, respectively, for the SLAM and 86% and 94%, respectively, for the SLEDAI). Serum C4d and Bb were more sensitive indicators of current moderate-to-severe lupus disease activity at all 3 visits than were serum C5b-9, C3, and C4. C3 and C4 were more specific indicators of moderate-to-severe disease activity. Serum C4d and Bb were more sensitive at predicting moderateto-severe disease activity at subsequent visits than were C5b-9, C3, and C4. Urine C3d was better than C3, plasma C4d, Bb, C5b-9 and anti-double-stranded DNA antibody in distinguishing patients with from those without acute lupus nephritis (P = 0.02).