A comprehensive analysis of available bioinformatic enrichment tools has recently been published. Based on the algorithm applied, the enrichment tools can be classified into three classes: singular enrichment analysis ; gene set enrichment analysis ;MS0015203 and modular enrichment analysis. In all tools, the input list of genes is mapped to the biological terms in databases, and then statistical analysis examines the enrichment of gene members for each of the annotation terms and corrects for multiple testing. We applied several SEA tools for the same input gene lists, and only enriched categories obtained with several tools were considered indicative of genuine prediction. This strategy, based on testing multiple tools, is recommended in order to obtain the most satisfactory results. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes are the two main annotation databases collecting biological knowledge of genes, which make them very suitable for bioinformatics scanning for enrichment analysis. Currently,AM4113 GO contains information for 18261 human gene products, while KEGG maps 373 different pathways. Our goal was to identify the functional categories that are consistently overrepresented in a statistically significant way in the list of differentially expressed genes inferred from the GEP studies on CRC prognosis. We first collected data from the 23 published independent GEP studies on prognosis of CRC to extract the genes reported in at least two of them, and then these genes were used for the systematic enrichment analysis with several independent SEA tools. This way, we overcame the lack of reproducibility observed in both the genes reported in individual GEP studies and the overrepresented categories reported by enrichment analysis tools, and could identify consistently enriched categories. Despite the variation in the number of overrepresented categories reported by the different enrichment tools, several categories were reported by many of the tools used.