Aim of the study: Infusions of Coreopsis tinctoria Nutt. flowering tops have been used traditionally in Portugal to control hyperglycaemia and a previous study revealed that daily administration of the infusion during a 3-week period promoted the recovery of glucose tolerance by a mechanism different from inhibition of glucose absorption and direct promotion of insulin secretion. We know report the study of the ethyl acetate fraction of Coreopsis tinctoria flowers infusion aiming to confirm flavonoids as bioactive metabolites. To give one step forward into the antihyperglycaemic mechanism of action of this traditionally used plant we also studied the activity of Coreopsis tinctoria flavonoids on the pancreatic function of glucose-intolerant rats. A standard antioxidant, Trolox, was also studied for comparative purposes as the antioxidant mechanism has been frequently purposed as one of the mechanisms mediating antihyperglycaemic effects of flavonoid-rich extracts. Material and methods: Thirteen compounds, mainly of flavanone and chalcone flavonoidal type, have been identified in this fraction by HPLC-DAD-ESI-MS/MS, and the major one (marein) quantified by HPLC-UV. The fraction (125 mg containing 20 mg of marein/kg b.w.) and Trolox (50 mg/kg b.w.) were administered daily by oral gavage to normal and STZ (40 mg/kg b.w.)-induced glucose-intolerant Wistar rats for 3 weeks. Blood glucose levels were measured weekly by Oral Glucose Tolerance Test. Pancreatic function was evaluated by plasma lipase of treated and non-treated glucose-tolerant and- intolerant rats after the 3-week treatment period. Results: After 2 weeks oral treatment with Coreopsis tinctoria AcOEt fraction the animals were no longer glucose-intolerant, an effect maintained over the remaining experimental period. Additionally, plasma lipase values of glucose-intolerant animals treated with the AcOEt fraction (13.5 +/- 0.84 U/L) showed a clear reduction when compared with the glucose-intolerant group (34.60 +/- 1.76 U/L: P<0.001) and normoglycaemic control (8.35 +/- 0.69 U/L) demonstrating recovery of pancreatic function. On the other hand, treatment with standard antioxidant Trolox had no effect on glucose homeostasis of glucose-intolerant rats. The oral treatment with Coreopsis tinctoria fraction caused no hepatotoxicity, as determined by blood alanine and aspartate transaminases, and had also no effect on glucose homeostasis and pancreatic function of normal rats. Conclusions: AcOEt fraction, containing the same amount of marein as the infusion, promoted glucose tolerance regain in the rats more quickly, which means that the bioactivity is probably due to the several flavonoids present in Coreopsis tinctoria extracts and not to marein alone. The results also strongly suggest that these compounds act by promoting pancreatic cell function recovery from STZ-induced injury, possibly through a mechanism of action other than merely antioxidant mediated.