Rheumatologists from 4 countries, representing 8 rheumatology centers, tested 3 systemic lupus erythematosus (SLE) disease activity indices: the SLE Disease Activity Index (SLEDAI) from Toronto; the Systemic Lupus Activity Measure (SLAM) from Boston and the British Isles Lupus Assessment Group (BILAG) for their reproducibility and validity in the assessment of real patients. Seven patients representing a spectrum of disease manifestations and activity were each examined by 4 of 7 observers from all centers except Toronto, using a Youden square design. Each observer completed all 3 indices and a category rating scale for disease activity on each of the 4 patients seen. All 3 indices detected differences among patients. There was no detectable observer effect among the 7 observers with each of the 3 indices. There was a detectable order effect with the SLAM. The 3 indices are comparable and reproducible for evaluating disease activity in SLE.