This paper presents a groupwise graph-theory-based approach to define nodes for network analysis in functional MRI (fMRI) data. The method ensures functional homogeneity within each subunit and consistency across multiple groups of healthy volunteers. The authors address the issue of selecting the appropriate number of nodes in the brain and demonstrate high reproducibility of parcellation results across different numbers of subunits (100, 200, and 300). The proposed approach is compared with existing methods, including group ICA and two other groupwise parcellation approaches, and is shown to outperform them in terms of classification accuracy and spatial continuity. The results are validated using synthetic and real resting-state fMRI data, and the authors provide a functional atlas with 100, 200, and 300 parcellations, along with tools to interface this atlas with various analysis packages. The study highlights the importance of functional homogeneity and reproducibility in defining nodes for network analysis in fMRI.This paper presents a groupwise graph-theory-based approach to define nodes for network analysis in functional MRI (fMRI) data. The method ensures functional homogeneity within each subunit and consistency across multiple groups of healthy volunteers. The authors address the issue of selecting the appropriate number of nodes in the brain and demonstrate high reproducibility of parcellation results across different numbers of subunits (100, 200, and 300). The proposed approach is compared with existing methods, including group ICA and two other groupwise parcellation approaches, and is shown to outperform them in terms of classification accuracy and spatial continuity. The results are validated using synthetic and real resting-state fMRI data, and the authors provide a functional atlas with 100, 200, and 300 parcellations, along with tools to interface this atlas with various analysis packages. The study highlights the importance of functional homogeneity and reproducibility in defining nodes for network analysis in fMRI.