The molecular pathogenesis of rheumatoid arthritis (RA) is still poorly understood and its clinical course can vary widely. We apply a systems biology approach to gain insight into the complex pathogenesis and disease heterogeneity. Genomics studies revealed considerable heterogeneity in global gene expression signatures between synovial tissue specimens from different patients with RA. Based on the molecular signatures, at least two distinct subsets of RA tissues could be identified. One class revealed abundant expression of gene clusters indicative of an ongoing activation of the adaptive immune response, whereas the other class resembled the expression pattern of osteoarthritic tissues, which are characterized by a low inflammatory gene expression signature and increased tissue remodeling. The molecular heterogeneity is featured not only at the whole synovial tissue level, but also at the level of fibroblast-like synoviocytes (FLS) cultured from those tissues. One of the most impressive features of our gene expression profiling studies is the clear correlation of the FLS phenotype with that of paired synovial tissue from which the cells were derived. One class of FLS is tightly related to the presence of lymphocytes in the lesions, whereas the other class of FLS suggests that synoviocyte-mediated invasion appears to be less dependent on infiltrating immune cells. These data support the notion that heterogeneity observed between synovial tissues is reflected in the FLS as a stable trait. Clearly, the list of genes that are differentially expressed between the tissue and FLS subgroups facilitates our understanding of the pathophysiology of the molecular forms of disease. Moreover, the differences in the gene expression profiles reflect important aspects of biological variation within the clinically diagnosed disease that provide a molecular basis for the well recognized but as yet poorly understood heterogeneity in RA and may help to (sub)classify rheumatic diseases.