Even with intensive insulin therapy it is impossible to reach physiological blood glucose levels in insulin-dependent diabetes mellitus. Because of the high costs and technical problems involved in islet cell transplantation broad applicability of this therapy seems uncertain. An alternative approach is the development of molecular-engineered insulin-producing clonal cell lines. The main interest is in rodent insulinoma cell lines and neuroendocrine AtT-20ins cells. This paper reviews the current knowledge about glucose-stimulated insulin secretion and the problems that have to be solved before these cells can be used for therapy in diabetes mellitus.