Serial immunological testing has been recently proposed for monitoring patients with lupus nephritis as routine serological tests have shown sub-optimal correlation with clinical status. To assess the value of urine cytology and urine sIL2R in the evaluation of patients with SLE, in particular those with lupus nephritis, we conducted a prospective double-blind study of 31 patients with SLE, during an 18-month period. A comparison of routine urinalysis with urine cytology and urine sIL2R was performed in 84 samples: 15 from patients without a history of renal involvement and 69 from patients with a history of renal involvement. A high urine cytology score (> or = 6), particularly in the presence of lymphoblasts, plasma cells or monocytes, was significantly associated with lupus nephritis in relapse. Urine sIL2R levels were significantly elevated during all SLE relapses, unrelated to the presence of renal involvement. Fifteen urine specimens were obtained at the time of a kidney biopsy: 9 with active lesions and 6 with inactive renal disease. UC score was 2.0 +/- 1.89 for those with absent activity, 8.4 +/- 3.4 for mild activity and 11.0 +/- 2.4 for moderate/severe activity (p < 0.001 between active vs inactive disease). No urinalysis parameter alone permitted distinguishing the degree of renal disease activity. In the subgroup of patients with renal disease urinalysis was overall less accurate than urine cytology or urinary sIL2R levels for predicting renal disease activity defined by biopsy. Urine cytology and urine sIL2R proved to be reliable measures of lupus activity.