GRETNA is a graph theoretical network analysis toolbox for imaging connectomics. It provides a comprehensive set of functions for analyzing brain networks using graph theory. The toolbox is based on MATLAB and is open-source, with a graphical user interface (GUI). It allows for the analysis of global and local network properties, and supports parallel computing. GRETNA enables statistical comparisons of network metrics and assessments of relationships between these metrics and clinical or behavioral variables. It also includes functionality for image preprocessing and network construction based on resting-state functional MRI (R-fMRI) data. After applying GRETNA to a publicly released R-fMRI dataset of 54 healthy young adults, it was demonstrated that human brain functional networks exhibit efficient small-world, assortative, hierarchical, and modular organizations. GRETNA is freely available on the NITRC website. The toolbox is designed to be flexible and efficient, allowing researchers to perform various network analyses, including network construction, analysis, and comparison. It supports different types of network analysis, including binary and weighted networks, and allows for the selection of different thresholding procedures. GRETNA is capable of performing parallel computing in the network construction and analysis modules, which can significantly shorten the duration of network analyses of large datasets. The toolbox is particularly useful for studies based on R-fMRI data and is compatible with various imaging modalities and species. GRETNA is a valuable tool for researchers in the field of brain connectomics, providing a comprehensive and flexible platform for analyzing brain networks.GRETNA is a graph theoretical network analysis toolbox for imaging connectomics. It provides a comprehensive set of functions for analyzing brain networks using graph theory. The toolbox is based on MATLAB and is open-source, with a graphical user interface (GUI). It allows for the analysis of global and local network properties, and supports parallel computing. GRETNA enables statistical comparisons of network metrics and assessments of relationships between these metrics and clinical or behavioral variables. It also includes functionality for image preprocessing and network construction based on resting-state functional MRI (R-fMRI) data. After applying GRETNA to a publicly released R-fMRI dataset of 54 healthy young adults, it was demonstrated that human brain functional networks exhibit efficient small-world, assortative, hierarchical, and modular organizations. GRETNA is freely available on the NITRC website. The toolbox is designed to be flexible and efficient, allowing researchers to perform various network analyses, including network construction, analysis, and comparison. It supports different types of network analysis, including binary and weighted networks, and allows for the selection of different thresholding procedures. GRETNA is capable of performing parallel computing in the network construction and analysis modules, which can significantly shorten the duration of network analyses of large datasets. The toolbox is particularly useful for studies based on R-fMRI data and is compatible with various imaging modalities and species. GRETNA is a valuable tool for researchers in the field of brain connectomics, providing a comprehensive and flexible platform for analyzing brain networks.