Abstract A significantly body of data on gene expression patterns in autoimmune diseases has been generated by microarray analysis. Although results are very promising, there are many factors that have detracted from the data. Indeed, no common methodological directions are available. Similarly, collection techniques, processing methods, and statistical approaches are often different. The impact of future studies will depend on the comparison of large datasets to validate results and must include rigorous statistical analysis. To better illustrate the issue we review herein the gene expression patterns observed in five representative autoimmune diseases, i.e. systemic lupus erythematosus, multiple sclerosis, rheumatoid arthritis, dermatomyositis, and primary biliary cirrhosis. We also emphasize how, once potential chromosome regions or pathways are identified, specific array design will be a powerful resource when used on large and representative populations.