OBJECTIVE To determine the clinical significance of serum lipoprotein(a) [Lp(a)] levels in patients with systemic lupus erythematosus (SLE). METHODS Serum Lp(a) levels in 77 patients with SLE were measured by turbidimetric immunoassay. RESULTS The median serum Lp(a) levels in all the SLE patients (14.4 mg/dl) and in those with active disease (Group A; 19.6 mg/dl, n = 39) at admission were significantly higher than those in healthy subjects (11.9 mg/dl, p < 0.05 and p < 0.001, respectively). The serum Lp(a) levels in SLE patients correlated directly with the serum cholesterol (p < 0.001) and urinary protein (p < 0.001) levels and inversely with the serum albumin levels (p < 0.02). Analysis limited to Group A patients with renal disease (Group A + RD, n = 28) revealed that the median serum Lp(a) level at the time of admission (OM) was significantly higher than those at 6 months before (-6M, p < 0.01) and at 6 months after admission (+6M, p < 0.01). Moreover, the serum Lp(a) level decrease from 0M to +6M in Group A+RD correlated significantly with the serum albumin level increase (p < 0.05). Multiple regression analysis demonstrated that the serum albumin level increment, the SLEDAI score decrement, the cholesterol level at 0M and the total dose of oral corticosteroids administered during the 0M to +6M period contributed independently and significantly to the serum Lp(a) level decrement from 0M to +6M in Group A + RD. CONCLUSIONS Our study is the first to reveal that hypoalbuminemia appearing during disease flare plays an important role in increasing the serum Lp(a) levels in lupus patients with renal disease and shows that corticosteroid treatment reduced the elevated serum Lp(a) levels almost to original levels.