Agricultural production must increase significantly to meet the needs of a growing global population with increasing per capita consumption of food, fiber, building materials, and fuel. Consumption already exceeds net primary production in many parts of the world.  In addition to reducing consumption, there are two options to meet these needs: production intensification and land conversion.  The largest gains achievable through agricultural intensification will likely occur on lands with the largest unrealized production potential, or yield gap. These lands have high potential production and low current production. Similarly, the highest returns on investments to be gained by land conversion should occur on lands with the highest potential production, assuming similar infrastructure, per acre conversion costs, and other market conditions.  This paper describes how a new cloud-based Land-Potential Knowledge System (LandPKS; www.landpotential. org) will allow land potential to be defined explicitly and dynamically for unique and constantly changing soil and climate conditions and to be updated based on new evidence about the success or failure of new management systems on different soils. The knowledge engine (figure 2), together with simple applications for mobile phones, will also facilitate more rapid and complete integration and dissemination of local and scientific knowledge about sustainable land management.