Disease Ontology (DO) enrichment analysis is an effective means to discover the associations between genes and diseases. However, most current DO-based enrichment methods were unable to solve the over enriched problem caused by the “true-path” rule. To address this problem, we presents EnrichDO, a double weighted iterative model, which is based on the latest annotations of the human genome with DO terms and integrates the DO graph topology on a global scale. On one hand, to reinforce the saliency of direct gene-DO annotations, different initial weights are assigned to directly annotated genes and indirectly annotated genes, respectively. On the other hand, to detect locally most significant node between the parent and its children, less significant nodes are dynamically down-weighted. EnrichDO exhibits high accuracy that it can identify more specific DO terms, which alleviates the over enriched problem.EnrichDO encompasses various statistical models and visualization schemes for discovering the associations between genes and diseases from biological big data. Currently uploaded to Bioconductor, EnrichDO aims to provide a more convenient and effective DO enrichment analysis tool.