Although there is little information from primary or secondary prevention trials on cholesterol-lowering medication in diabetic patients, the reduction of elevated cholesterol is widely recommended for this group. The American Diabetes Association (ADA) recommends drug therapy in diabetic patients if low density lipoprotein (LDL)-cholesterol remains at > 130 mg/dl, or > 100 mg/dl in patients with macroangiopathy, after dietary intervention. When cholesterollowering medication is indicated, the choice of the drug must take into account the other lipid abnormalities that are often present and the need to maintain optimal glycaemic control. In the present study we compared the efficacy and safety of the novel HMG-CoA reductase inhibitor atorvastatin at the dose of 10 mg/day with simvastatin , lovastatin and pravastatin at doses of 10, 20 and 20 mg/day, respectively, and placebo, in type 2 diabetic patients with moderate elevation of LDL-cholesterol with or without elevation of triglycerides. All the quoted agents are enzyme inhibitors effective in lowering LDL-cholesterol in humans. The efficacy endpoints were the mean per cent changes in plasma LDL-cholesterol (primary), total cholesterol, triglycerides, and high-density lipoprotein (HDL)-cholesterol concentrations from baseline to the end of treatment (24 weeks). Atorvastatin at a dose of 10 mg/day produced: (1) a significant reduction in LDL-cholesterol (-37%) in comparison with equivalent doses of simvastatin (-26%), pravastatin (-23%), lovastatin (-21%), and placebo (-1%); (2) HDL-cholesterol increases (7.4%) comparable to or greater than those obtained with simvastatin (7.1%), pravastatin (3.2%), lovastatin (7.21%), and placebo (-0.5%); (3) a significantly greater reduction in total cholesterol (- 29%) than that obtained with simvastatin (-21%), pravastain (-16%), lovastatin (-18%), and placebo (1%); and (4) a significantly greater reduction in triglycerides than that obtained with all the other drugs and placebo. In all treatment groups no significant variation in fibrinogen concentration was observed. All reductase inhibitors studied had similar levels of tolerance. There were no incidents of persistent elevations of serum aminotransferases or myositis.