Since nutrition-sensitive feedback signals normally act to maintain relatively stable levels of both available and stored nutritional resources, failure in one or more of these feedback signals could plausibly lead to obese phenotypes. The glucostatic hypothesis in its original form posited that glucose serves as a physiological satiety factor (in the sense that post-prandial increases in plasma glucose cause meal termination), but in this form the hypothesis has been difficult to prove, and, especially since the discovery of leptin, the glucostatic hypothesis has largely been abandoned. Nevertheless, reduction of plasma glucose levels or glucose signaling produces a profile of neuroendocrine responses similar to those produced by leptin deficiency. Since leptin is not a physiological satiety factor (because it does not increase before meal termination), yet leptin deficiency causes obesity, we suggest that the glucostatic hypothesis be re-formulated without reference to satiety (i.e., short-term effects on food intake). Instead we argue that like leptin signaling, glucose signaling regulates long-term energy balance, in part by regulating metabolic rate but also by chronically regulating food intake. We further speculate that high-fat diets produce obesity in part because carbohydrates are, per calorie, more effective than lipids to reduce food intake and increase metabolic rate. In support of this glucoadipostatic hypothesis, the present review examines evidence that obesity and the metabolic syndrome may be due to reduction in neuroendocrine sensitivity to glucose leading to increased metabolic efficiency.