Little information exists to quantify the functional status and economic consequences of lipid-lowering therapy in elderly patients. We describe the design of the cost-effectiveness component of the first large, randomized, placebo-controlled trial of lipid-lowering therapy in subjects aged 70 years or older. The PROspective Study of Pravastatin in the Elderly at Risk has randomized 5804 men and women 70-82 years old, with existing vascular disease or related risk factors, to receive 40 mg/d of pravastatin or placebo. The cost-effectiveness study will be based on within-trial observations of differences between the two study arms in rates of myocardial infarction, stroke and related vascular disease outcomes (including vascular dementia). In addition to comparing within-trial clinical outcomes, we will model the projected changes in life expectancy and cardiovascular outcome rates that would be seen with lifelong use of the drug, based on the risk reduction rates seen in the trial as well as baseline observational data from the three countries where the trial is being conducted (Scotland, Ireland and the Netherlands). A state transition (Markov) model will be constructed to estimate the likely states of health, functional status and health care resource utilization (including lipid-lowering drug costs) in a cohort of elderly patients with versus without pravastatin therapy over a series of 1-year cycles until death. In addition, a standard measure of utility (the Health Utilities Index) will be administered to all study subjects to permit calculation of quality-adjusted life-years gained with this regimen. This approach will make it possible to go beyond the calculation of a single endpoint for each subject, and to translate the trial findings into definitions of effectiveness and outcomes that will be relevant to the ongoing debate concerning how best to relate the benefits of such medications to their costs.