The paper introduces eggNOG-mapper, a tool for fast genome-wide functional annotation using orthology assignments. Orthology-based functional inference is more accurate than homology-based methods, but it is computationally intensive and less accessible. eggNOG-mapper leverages precomputed clusters and phylogenies from the eggNOG database to perform orthology assignments efficiently. The tool was benchmarked against BLAST and InterProScan, showing that it reduces false positive assignments by 11% and increases the ratio of experimentally validated terms by 15%. Compared to InterProScan, eggNOG-mapper achieved similar proteome coverage and precision while predicting 41 more terms per protein and increasing the rate of experimentally validated terms by 35%. eggNOG-mapper outperformed both tools in the CAFA2 benchmark and metagenomics data annotation, with faster computation times. The tool is available as a standalone package and an online service, facilitating functional annotation for researchers.The paper introduces eggNOG-mapper, a tool for fast genome-wide functional annotation using orthology assignments. Orthology-based functional inference is more accurate than homology-based methods, but it is computationally intensive and less accessible. eggNOG-mapper leverages precomputed clusters and phylogenies from the eggNOG database to perform orthology assignments efficiently. The tool was benchmarked against BLAST and InterProScan, showing that it reduces false positive assignments by 11% and increases the ratio of experimentally validated terms by 15%. Compared to InterProScan, eggNOG-mapper achieved similar proteome coverage and precision while predicting 41 more terms per protein and increasing the rate of experimentally validated terms by 35%. eggNOG-mapper outperformed both tools in the CAFA2 benchmark and metagenomics data annotation, with faster computation times. The tool is available as a standalone package and an online service, facilitating functional annotation for researchers.