Abstract A 34% reduction in the incidence of definite coronary heart disease events was observed in dyslipidemic men treated with gemfibrozil in the Helsinki Heart Study, a controlled 5-year, double-blind primary prevention trial for coronary heart disease. Over the entire study period, gemfibrozil therapy induced mean decreases of 10% in serum total cholesterol levels, 11% in low-density lipoprotein (LDL) cholesterol, 35% in triglyceride levels, and a mean increase of 11% in high-density lipoprotein (HDL) cholesterol level, compared with placebo. The differences in percentage changes in LDL cholesterol between gemfibrozil- and placebo-treated men varied among Fredrickson hyperlipoproteinemia types; after 1 year of treatment the difference was greatest for type IIA hyperlipoproteinemia (14 percentage units) and smallest for IIB hyperlipoproteinemia (3 percentage units). The treatment-associated changes in HDL cholesterol and triglycerides did not differ materially between the 3 hyperlipoproteinemia types, when calculated in the same way. The gemfibrozil-associated reduction in incidence of definite coronary events varied among Fredrickson types and among tertiles of baseline HDL cholesterol and triglycerides. The greatest rate reductions were seen in subjects with type IIB hyperlipoproteinemia, low initial HDL level or high initial triglycerides. These results suggest that subjects with low HDL cholesterol and type IIB hyperlipoproteinemia (and possibly type IV hyperlipoproteinemia) would benefit from treatment with gemfibrozil.