Twelve elderly non-insulin dependent diabetic patients took part in a double-blind, cross-over, randomized study comparing simvastatin 30 mg/day and placebo. Each treatment period lasted 3 weeks and was separated by a 3 week wash-out period. At the end of each treatment period all subjects underwent in randomized order an oral glucose tolerance test (OGTT; 75 g) and an euglycaemic hyperinsulinaemic (50 mU/kg.h) glucose clamp. Simvastatin compared to placebo significantly reduced plasma total cholesterol (7.9 vs 5.3 mmol.l-1), LDL-cholesterol (7.2 vs 4.3 mmol.l-1), triglycerides (2.9 vs 2.1 mmol.l-1), free fatty acids (1106 vs 818 mmol-1) and glucose (7.4 vs 6.6 mmol.l-1) levels. After simvastatin, and in the last 60 min of the glucose clamp, there was an improvement in the action of insulin as demonstrated by stronger inhibition of hepatic glucose output (2.7 vs 5.2 mumol.kg-1.min-1) and stimulation both of the glucose disappearance rate (26.3 vs 19.5 mumol.kg-1.min-1) and glucose metabolic clearance rate (4.3 vs 3.6 ml.kg-1.min-1). The changes in glucose turnover parameters were significantly correlated with basal plasma free fatty acids and were independent of plasma glucoregulatory hormones. In conclusion, simvastatin seems to exert beneficial effects both on lipid and glucose metabolism.