New clinical scales for semiquantitating disease activity in systemic lupus erythematosus (SLE) are widely used in research. They are reliable and valid measures. One of the original scales, the Systemic Lupus Activity Measure (SLAM), has been modified based on experience with it in multi-observer studies and training of individuals in its use. We tested the psychometric properties of the revised SLAM (SLAM-R). SLAM-R was tested on 30 SLE patients, who fulfilled 1997 revised ACR criteria and were selected to represent a range of disease activity. The patients were evaluated independently by two physicians, who studied the instruction booklet and who had never used SLAM-R, on two occasions 2-4 weeks apart. At the first visit, the physician's global assessment of activity using visual analog scale, anti-dsDNA Ab, C3 and C4 were checked for construct validity. The psychometric properties were analyzed with nested analysis of variance and Pearson's correlation coefficient using SAS. All patients were female, the median age was 31 (15-52) y, and the mean score of SLAM-R was 10.5 +/- 5.3 (3-26). Estimates of reliability were 0.78 of inter-rater, 0.61 of inter-visit, 0.76 of physician 1 between visits, and 0.56 of physician 2 between visits. Among subcategories except 'Eye,' the 'Gastrointestinal' category had the highest (0.96) and the 'Neuromotor' category had the lowest inter-rater reliability (0.50). With respect to construct validity, the correlation of SLAM-R scores with the disease activity variables except C4 was high and statistically significant. In conclusion, the SLAM-R is reliable and valid for measuring clinical disease activity in SLE.