Serial analysis of gene expression (SAGE) allows a comprehensive profiling of gene expression within a given tissue and also an assessment of transcript abundance. We generated SAGE libraries from normal and neoplastic plasma cells to identify genes differentially expressed in multiple myeloma (MM). Normal plasma cells were obtained from palatine tonsils and MM SAGE library was generated from bone marrow plasma cells of MM patients. We obtained 29,918 SAGE tags from normal and 10,340 tags from tumor libraries. Computer-generated genomic analysis identified 46 upregulated genes in the MM library. Ten upregulated genes were selected for further investigation. Differential expression was validated by quantitative real-time PCR in purified plasma cells of 31 patients and three controls. P53CSV, DDX5, MAPKAPK2 and RANBP2 were found to be upregulated in at least 50% of the MM cases tested. All of them were also found upregulated in MM when compared to normal plasma cells in a meta-analysis using ONCOMINE microarray database. Antibodies specific to DDX5, RANBP2 and MAPKAPK2 were used in a TMA containing 57 MM cases and confirmed the expression of these proteins in 74%, 96%, and 21% of the MM samples, respectively. Analysis of differential expression using SAGE could identify genes important for myeloma tumorigenesis (P53CSV, DDX5, MAPKPK2 and RANBP2) and that could potentially be useful as therapeutic targets.