The recent rapid increase in the prevalence of obesity across the world is undoubtedly due to changes in diet and lifestyle. However, it is also indisputable that different people react differently to this change in environment and this variation in response is likely to be genetically determined. While for the majority of people this effect is presumed to be polygenic in origin, there is now strong evidence for a small number of genes having a large effect in some families with severe obesity. Studies of these families, coupled with parallel studies in murine models, have provided novel insights into the molecules involved in the regulation of appetite, energy expenditure and nutrient partitioning. We review here the lessons we have learnt from mouse models of obesity, both naturally occurring and artificially generated through targeted gene deletions, and more importantly from human monogenic syndromes of obesity. These have illuminated the critical role in which the central leptin melanocortin pathway plays in the control of mammalian food intake and body weight.