The competent interaction of humans with their environment is essential for its well-being and therefore also for the well-being of humans. Such behavior is characterized by its relatedness to the environment. To behave in an environmentally competent manner, human individuals must therefore learn that they and their actions are part of a human-environment relationship. Humans can do this by learning to associate themselves and their actions with their environment. Learning this association between themselves and the environment has the potential to establish the motivation for environmentally competent behavior in human individuals. For this type of learning, human individuals must be enabled to experience that their behavior results in changes that simultaneously affect their environment and themselves. This can be implemented well in immersive virtual reality. We have therefore developed EcoWalk. This is a life-sized three-dimensional virtual environment for associative learning by walking on the omnidirectional treadmill of Infinadeck. The virtual environment of EcoWalk is programmed in a manner that depending on the carbon footprint of food its purchase is either triggering or reversing climatic environmental changes such as heavy rain or flooding. We are currently investigating scientifically, whether EcoWalk can serve to foster the learning of eco-friendly behavior and how a low-cost version of EcoWalk can be used in daily practice.