Plants have been central to human life across evolutionary time as sources of food and raw materials for a variety of uses (e.g., artifact construction). However, plants also manufacture toxic chemicals and mechanical defenses to protect themselves from herbivores. These circumstances create a fundamental problem: How does each individual human learn which plants in her local environment are food (or otherwise useful) and which plants are fatal? Because there are no morphological features of plants that reliably signal human-relevant edibility or toxicity, employing a trial-and-error strategy to learn about the specific plants in an environment would be extremely costly. Instead, I argue that human cognitive architecture contains social learning mechanisms specialized for acquiring information about plants over the course of ontogeny, called Plant Learning and Avoiding Natural Toxins, or PLANT. In this talk, I will present empirical evidence for PLANT systems from studies of infants and young children across several cultures, as well as comparative studies of nonhuman primate species. From a theoretical perspective, PLANT systems provide a paradigmatic case of the ways in which evolutionary processes can build social learning mechanisms tailored to the affordances of specific kinds of content. From a practical perspective, understanding how plants have shaped the human mind provides insight into why children tend to avoid eating a diet rich in fruits and vegetables, and why adults can have blind spots around issues of biodiversity. A research program centered on PLANT systems can (1) provide important insights into the cognitive systems that facilitate human interactions with the environment and (2) develop real-world programs and interventions that promote environmentalism and sustainable practices in infancy and early childhood.