In a double-blind, randomized crossover study, 29 patients with non-insulin-dependent diabetes mellitus (NIDDM) and hyperlipoproteinemia were treated with gemfibrozil (1,200 mg/d) or simvastatin (10 mg/d) for 4 months. After gemfibrozil treatment, the insulin concentration was increased during the major part of the intravenous glucose tolerance test (IVGTT) and during the hyperinsulinemic euglycemic clamp. Similar but less pronounced elevations were caused by simvastatin. Insulin sensitivity decreased by 27% and 28% during gemfibrozil and simvastatin treatment, respectively. Low-density lipoprotein (LDL) cholesterol was decreased with simvastatin treatment by 24%. The LDL cholesterol level was not changed by gemfibrozil, but very-low-density lipoprotein (VLDL) cholesterol was reduced by 40%. The VLDL triglyceride concentration was reduced to a significantly greater extent by gemfibrozil. After gemfibrozil treatment, lipoprotein(a) [Lp(a)] was decreased by 24%, and the plasma free fatty acid (FFA) concentration was increased by 20% and skeletal muscle lipoprotein lipase activity (LPLA) by 37%. Although simvastatin more effectively decreased LDL cholesterol levels and the LDL to high-density lipoprotein (HDL) ratio, it cannot be claimed unreservedly that this drug is necessarily preferable in NIDDM patients. Gemfibrozil improved triglyceride removal and decreased VLDL concentrations, with qualitative changes in LDL. The apparent effects on insulin sensitivity are difficult to evaluate and need further study.