The control of glucose metabolism by pancreatic endocrine cells throughout life relies on a tight regulation of the mass of insulin-producing beta-cells. How this homoeostasis is achieved is not well understood. Over the last few years, experimental rodent models with altered beta-cell mass, and, more recently, new transgenic approaches designed to tackle this problem, have provided abundant information. Processes such as beta-cell proliferation and apoptosis, or even beta-cell differentiation from poorly characterized progenitor cells, whether immature or differentiated, appear to be implicated. A complex picture is thus emerging in which the nature of the pancreatic lesion appears to determine the kind of regenerative response. The environment formed by acinar and ductal cells, and also by vascular and neuronal structures, which surround islets and penetrate into their beta-cell core, might play crucial roles so far unsuspected, which should be explored in the near future.