The brain is a highly complex system that is organized by large-scale connectomes. In the sensory brain, specific connectomes, i.e., intrinsic neural networks (ICNs), have been well-established for vision, audition, and somatosensation. However, the definition of the chemosensory connectome has lagged. Leveraging the large and high-quality resting-state functional magnetic resonance imaging (rs-fMRI) dataset of nearly 900 participants from the Human Connectome Project (HCP), we combined resting-stating functional connectivity analysis with graph-theoretical analysis to identify the human olfactory connectome/network (Arnold et al., 2020). Our study demonstrated that (1) the olfactory neural network consists of cortical and subcortical regions widely distributed in the frontal and temporal lobes; (2) the olfactory network is organized by three subdivisions—the sensory, limbic, and frontal subnetworks; and (3) the olfactory network resembles a highly efficient system characterized by a high degree of global integration balanced with a high degree of local segregation (i.e., circuit specialization). Highlighting its reliability and generalizability, the composition of the olfactory network mapped closely onto one that was extracted from our independent rs-fMRI dataset. Moreover, the degree of local segregation positively predicted olfactory discrimination performance in our sample. The composition and topological organization of the human olfactory network would illuminate the machinery underlying olfaction, the most intriguing sense for its heterogeneous and yet harmonized functions in sensory perception, emotion, neuroendocrine, and homeostasis. Therefore, we applied this olfactory network composition and topological principles to fMRI responses to odor and sweat stimuli of aversive and neutral emotion. Findings from this new study provide important connectome-level insights into the functional architecture underlying this uniquely multipurpose system of olfaction.