During the past 30 years great effort has been put into establishing an insulin-secreting beta cell line that retains normal regulation of insulin secretion, but only few of these attempts have been successful. To overcome the limited availability of primary beta cells and to include the principles of the 3Rs into the field of diabetes mellitus research, numerous investigators used X-rays or viruses to induce insulinomas, in vitro transformation, derivation of cells from transgenic mice or even non-islet cells to produce immortalised beta cell lines. The most widely used insulin-secreting cell lines are RIN, HIT, MIN, INS-1 and TC cells. These cells produce insulin and small amounts of glucagon and somatostatin. Some of them are only poorly responsive to glucose, others respond to glucose well, but their concentration-dependence curve is markedly shifted to higher sensitivity. Despite problems associated with beta cell cultures, these cell lines have provided some valuable information about physiological processes. However, an urgent need to establish a "normal" beta cell line of human or pig origin remains.