MD) is a simple process that utilizes heat to drive vapor to pass through a porous hydrophobic membrane and obtains clean and fresh water in the permeate, while liquid water is blocked by surface hydrophobicity. [2] The potential of using lowgrade heat has made MD a promising low-cost and sustainable water desalination technology. [3] MD is also a key technology in many zero-discharge sustainable processes because the "last-mile" of such processes often involves a high-salinity water solution that is failed to be treated by other desalination technologies, while MD can in principle harvest 100% water from either polluted or high salinity waters. [4] The reported MD membranes are typically prepared from hydrophobic polymers such as polypropylene (PP), poly(vinylidene fluoride) (PVDF), and poly(tetrafluoroethylene) (PTFE). [5] However, these membranes have suffered from low flux. Structural optimization using, for example, sandwiched, hierarchical, or Janus structures, has been studied extensively to improve the flux. [6] Nevertheless, the best-reported membrane flux thus far is less than 80 L m −2 h −1 (LMH) in the direct contact membrane distillation (DCMD) mode even under a high temperature gradient of 90/20 °C. [2a,6a,7] As a result, applications of these membranes in sustainable processes such as solar-driven MD systems have shown far below the practically required water capacity.In our previous work, we found a graphitic carbon nanowire membrane showing a superior water flux of 400 LMH at 90/20 °C in vacuum membrane distillation process (VMD) mode because of the fast water transport on the graphitic surface and the short transport pathway. [8] However, the 1D forest-like carbon nanowires also caused profound concentration polarization. In contrast, the atomically thin 2D graphene, another graphitic carbon material, with a similar surface masstransfer property to the 1D carbon nanowires/tubes, could ameliorate these challenges, while holds potentially even shorter transport distance thus higher MD performance. [9] However, because of its impermeability and insufficient strength over large areas, a functional graphene membrane will require postsynthesis pore generation and transfer to porous support. Both have been proved to be of great challenge in the current studies. [9c,10] Membrane distillation has attracted great attention in the development of sustainable desalination and zero-discharge processes because of its possibility of recovering 100% water and the potential for integration with low-grade heat, such as solar energy. However, the conventional membrane structures and materials afford limited flux thus obstructing its practical application. Here, ultrathin nanoporous graphene membranes are reported by selectively forming thin graphene layers on the top edges of a highly porous anodic alumina oxide support, which creates short and fast transport pathways for water vapor but not liquid. The process avoids the challenging pore-generation and substrate-transfer processes required to prepare regu...