Whether hypoglycaemic sulphonylureas promote pancreatic beta-cell replication (beta-cytotrophic action) remains controversial largely because previous studies were based on semi-quantitative techniques only. Here, we used strict quantitative and reliable morphological methods to assess whether glibenclamide affects the replication rate and the mass of beta-cells in normal mice. The animals received a daily intraperitoneal injection of glibenclamide (2 mg/kg body weight) for 7-30 days. Two hours before being killed, they also received an injection of 5-bromo-2'-deoxyuridine, which accumulates in the nucleus of cells in the S-phase of the cell division cycle. Sections of the pancreas were then stained for immunodetection of insulin-containing beta-cells and replicating beta-cells (labelled nucleus). The treatment caused a degranulation of beta-cells, an increase in the percentage of beta-cells with a labelled nucleus, but no beta-cell hypertrophy. This stimulation of replication led to a small increase in the beta-cell mass at the end of the treatment. However, the effect of glibenclamide on beta-cell replication was short-lasting (less than 15 days) and was observed only when the treatment was started in young animals (23 or 30 days old, but not 42 days old). In conclusion, glibenclamide has a beta-cytotrophic effect which is transient and only seen in young animals. It is unlikely, therefore, that this effect plays a role in the treatment of adult diabetic patients.