Macroscopic functional connectomic analyses have identified sets of densely connected regions in the human brain, known as connectome hubs, which play a vital role in understanding network communication, cognitive processing, and brain disorders. However, anatomical locations of functional connectome hubs are largely inconsistent and less reproducible among extant reports, partly due to inadequate sample size and differences in image processing and network analysis. Moreover, the genetic signatures underlying the robust connectome hubs remain unknown. Here, we conduct the first worldwide voxelwise meta-connectomic analysis by pooling resting-state functional MRI data of 5,212 healthy young adults across 61 independent international cohorts with harmonized image processing and network analysis protocols. We identify highly consistent and reproducible functional connectome hubs that are spatially distributed in multiple heteromodal and unimodal regions, with the most robust findings mainly located in lateral parietal regions. These connectome hubs show unique, heterogeneous connectivity profiles and are critical for both intra- and inter-network communications. Using transcriptome data from the Allen Human Brain Atlas and BrainSpan Atlas as well as machine learning, we demonstrate that these robust hubs are significantly associated with a transcriptomic pattern dominated by genes involved in the neuropeptide signaling pathway, neurodevelopmental processes, and cellular metabolic processes. This pattern represents microstructural and metabolic substrates underlying the development and functioning of brain hubs. Together, these results highlight robustness of macroscopic connectome hubs of the human brain and their potential cellular and molecular underpinnings and have implications for understanding how brain hubs support the connectome organization in health and disease.