The global food system is essential for sustaining human life, yet in its current form it does so at a great cost to both human and environmental health. Food insecurity and diet-related disease are responsible for 70% of deaths worldwide, and the global extinction rate is ~1000 times estimated background rates due to the widespread destruction of wild habitat for agriculture. The production of livestock for animal protein is a key driver of both trends. This thesis considers the potential of edible insects in existing agricultural systems to mitigate these problems, with a focus on the role of chitoumou (Cirina butyrospermi), the commonly consumed shea caterpillar harvested from agroforestry systems in Burkina Faso, West Africa. I begin this thesis with a review of the role of edible insects in agricultural systems worldwide. Though certain insects are currently touted for commercial scale production, many popular edible insects are already harvested from existing systems, in which they prey on plant crops. These insects often cause damage to crop yields, but also provide vital provisioning, regulating, maintaining and cultural ecosystem services. Yet, few data are available to enable farmers and policymakers to weigh up the costs and benefits of edible insects in agricultural systems. Therefore, I examine the contributions of chitoumou – a popular edible insect in a region with acute food insecurity and environmental degradation – to food security, crop yields and biodiversity. Firstly, I evaluate the contribution of shea caterpillars to food security. I show that animal protein consumption and food security are higher during caterpillar season, and that higher food security is associated with caterpillar collection, sale and consumption. Shea caterpillars contribute positively to food security, but this effect is seasonal. Secondly, I investigate the relationship between defoliation by shea caterpillars and crop yields. Observational data show that defoliation does not have a negative association with yields of either shea (Vitellaria paradoxa) or maize (Zea mays). This challenges assumptions held by some stakeholders that it is necessary to take measures to eradicate these insects. Thirdly, I look at the relationship of caterpillars to biodiversity, using defoliation as a proxy for caterpillar abundance and bird abundance as a proxy for biodiversity. I find no significant relationship, suggesting that defoliation by caterpillars is not a significant threat to biodiversity in this region. Consequently, policymakers and smallholder farmers alike should recognize that shea caterpillars are not pests, and that the retention of shea trees will promote food security. However, policymakers in particular should be aware that the contribution of shea caterpillars to food security is seasonally limited. For this reason, assessments of food insecurity and strategies to mitigate food insecurity in this region are best conducted out of caterpillar season. Future research that aims to tackle acute global problems of human and environmental health using edible insects should consider the role of insects in existing agricultural systems, and quantify how such insects affect regions where these problems are most severe.