The hypoglycaemic effects of two quinolizidine alkaloids: lupanine and 2-thionosparteine were examined in non-diabetic and in streptozotocin-induced diabetic rats. The model of experimental diabetes can be considered to be related to diabetes mellitus type 2 with regards to the impairment of beta-cells' secretory function. A single intraperitoneal injection of 2-thionosparteine at a dose of 8.6 mg/kg lowered the blood glucose levels in diabetic rats at 90 and 120 min after administration and showed similar hypoglycaemic effects to glibenclamide and sparteine, which were used as reference substances. In contrast to glibenclamide, 2-thionosparteine did not result in a significant increase in plasma insulin levels in diabetic rats; an increase was only observed in the non-diabetic group. It was found that lupanine did not exert hypoglycaemic potency in diabetic and in non-diabetic animals and did not significantly increase plasma insulin concentration independent of the group examined. From this study we can state that 2-thionosparteine, but not lupanine, is confirmed to be a possible plasma glucose lowering agent. It is possible that 2-thionosparteine-dependent decrease in blood glucose level is not the only result of this drug's related insulin secretion.