Understanding the organisation and function of what Santiago Ramón y Cajal referred to as the impenetrable jungle of the brain has been one of the biggest fascinations of ancient and modern science. How does this exquisitely complex structure of close to 100 billion neurons form the basis of who we are, coordinating the processes such as our internal organ function, as well as directing our behaviour, perception and emotion? As a step closer to answering these questions, the work in this thesis aims to unravel the role of the brain in one of the most fundamental processes – maintenance of glucose homeostasis. The brain is constantly informed about the state of our internal environment and the external world; if a disruption occurs, it is able to coordinate physiological and behavioural responses to counterbalance it. This concept holds true for glucose as well. However, interestingly, here the brain not only receives inputs about the level of glucose in the blood, it also has the capability to measure – sense – glucose itself. I hypothesised that blood glucose homeostasis is exerted by glucose-responsive neurocircuitry within the brain, with discrete nodes responding to a rise or fall in glucose. The overarching aim of this thesis was to use mouse models to evaluate the identity and role of these circuits in controlling physiological and behavioural responses to blood glucose changes. I first examined a group of neurons located on the floor of the brain, that we hypothesised might have this exact function: the glucokinase (Gck) -expressing agouti-related protein (AgRP) neurons. The enzyme Gck is well-known for its crucial role within the pancreas, where it enables sensing of glucose levels and leads to glucose-stimulated insulin release. Here, we demonstrate its importance for AgRP neuronal function by genetically removing it from these cells in mice. We show that perturbing the function of this strikingly small population of approximately 3,000 neurons has profound impact on whole-body glucose homeostasis, causing insulin resistance, glucose intolerance, and a reduction in glucose-stimulated insulin secretion. Strikingly, we show that this effect is only seen in female mice, with males exhibiting no disturbances whatsoever. Next, I took a step back and observed the coordinated activity of the neural circuits across the entire brain induced by changes in blood glucose, from a drop (hypoglycaemia) to a rise (hyperglycaemia) in glucose levels. Here, I combined two powerful contemporary techniques, glucose clamps and tissue clearing, to (1) precisely manipulate blood glucose levels causing brain activation, and (2) subsequently render these brains transparent and scan them using light sheet microscopy. We create the first-of-its kind 3D atlas of the whole-brain glucose-responsive activity with a single-cell resolution, and we identify at least a dozen brain regions – potential nodes in the whole-brain glucose-responsive circuitry – showing discrete activity driven by blood glucose changes. Finally, we start to build a foundational framework to investigate the exact function of each of these nodes. Finally, I began to develop and adapt two novel technical approaches to quantify behavioural responses to low blood glucose (hypoglycaemia) in mice. In addition to the more well-known detrimental effects of high blood glucose and diabetes, hypoglycaemia is perhaps even more acutely dangerous as a constant supply of blood glucose is crucial for the brain function and survival. If fully validated, these methods could be used to advance our understanding of the brain circuits engaging protective mechanisms which defend against hypoglycaemia. I hope that this body of work will serve as a strong foundation for future mechanistic studies examining how blood glucose is maintained, how this process goes awry in disease such as diabetes, and answering whether we can specifically target brain neural circuits to find new drugs to treat, or even revert, diabetes.