BACKGROUND: A low serum level of high-density lipoprotein (HDL)-cholesterol is an independent risk factor for coronary heart disease (CHD). Fibrates, particularly gemfibrozil, have been shown to raise HDL-cholesterol levels and reduce the incidence of CHD. The literature on fibrate cost effectiveness is quite limited. OBJECTIVE: The objective of this analysis is to determine the cost effectiveness of the fibrates gemfibrozil and fenofibrate in the primary prevention of CHD. The target population includes patients with low levels of HDL-cholesterol, but without pre-existing CHD or other CHD risk factors sufficiently elevated to indicate drug therapy. STUDY DESIGN AND METHODS: From a societal perspective, a lifetime incremental cost-effectiveness model was developed to calculate baseline and treatment costs, life-years gained and QALYs gained. Model parameter values were taken from existing literature. In this ‚backward induction‘ model, the expected costs and outcomes for each 5-year time-interval are utilised in subsequent 5-year time period calculations over the patient’s entire lifetime. The study population consisted of a hypothetical cohort of males and females in the US aged 45-74 years, with low levels of HDL-cholesterol and no prior history of CHD. The base-case CHD risk factors for this population were obtained from the VA-HIT (Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial) population baseline characteristics, but assuming no prior CHD history. Estimates for the reduction in CHD risk associated with fibrate therapy reduction are also taken from the VA-HIT study. RESULTS: Using a societal cost-effectiveness threshold of US$50, 000 per QALY, primary prevention of CHD in patients with low HDL-cholesterol levels using generic gemfibrozil therapy is cost effective for all age and sex categories, in contrast to fenofibrate therapy, which is cost effective for males, but not for females at baseline risks levels. In the base-case scenario, because of their higher CHD lifetime risk, it is more cost effective to treat males than females with either gemfibrozil or fenofibrate. For males and females the cost per QALY decreases with age for most age intervals. Gemfibrozil is more cost effective than fenofibrate for all age-sex categories because of the assumed equal efficacy and the higher fenofibrate drug cost. In the comparison scenario, generic lovastatin was more cost effective than gemfibrozil for men except at age 45 years and women at all ages, and more cost effective than fenofibrate for both men and women. CONCLUSIONS: This analysis suggests that fibrate therapy, particularly with generic gemfibrozil, is cost effective in the primary prevention of CHD in individuals with low HDL-cholesterol levels, with or without elevated triglyceride levels. Certain patient subgroups, such as those with elevated triglyceride levels, smokers and those with diabetes mellitus are likely to achieve both CHD risk reduction and overall savings in net expected medical care costs. Comparable cost-effectiveness results are also shown for lovastatin therapy in the target patient population. Gemfibrozil dominates fenofibrate because of the lower cost of therapy (direct and indirect costs). These conclusions are robust to reasonable changes in model parameter values.