Microarray and Next-Gen Sequencing technologies have generated a huge amount of genome scale expression data. Many software tools permit extraction of expression profiles/data for a given gene or a set of genes. It has been established that genes with similar expression profiles are likely to be associated functionally. Unfortunately, there has been no simple online application for mining expression databases for genes with similar profiles. We have developed a method and an associated web tool - Microarray Meta-Miner (http://exon.niaid.nih.gov/MMM/) - to identify genes with expression profiles similar to that of the query gene. Using the microarray meta-data from the ATLAS gene expression database and the NIH Biowulf cluster computing facility, we computed eight different vector similarity metrics (Pearson/Spearman/Kendall correlation coefficients, mutual information, chi-square, Euclidean distance, purity, and cosine similarity) for every gene's expression profile against every other gene's profile. Top scoring hits from the individual metrics were integrated and scored for overlap, generating a matrix of similar expressions. The MMM web tool returns the list of similarly expressed genes for the users query, along with links to annotations, individual expression profiles, and all expression profiles. MMM also retrieves and displays the known interaction data between the similarly expressed genes from the STRING interaction database. The experiment distribution information for the set of similarly expressed genes is also displayed. The current version of MMM supports only human data. Future plans include adding support for other organisms.