OBJECTIVE To evaluate the cost effectiveness from a UK health and social services perspective of antiplatelet therapies tested in the Second European Stroke Prevention Study (ESPS-2) in preventing recurrent stroke. To demonstrate the value of modelling studies in this area. METHODS A decision-analytic model was developed to evaluate health outcomes and associated costs. Sources of data for efficacy, adverse events, background event risks, disability and mortality were ESPS-2, the Oxfordshire Community Stroke Project and UK national statistics. Published national unit costs were applied to clinician panel estimates of resource use for acute stroke, rehabilitation and long term care. Outcome measures were strokes or disabled life-years averted, and disability-free, stroke-free or quality-adjusted life-years gained. METHODS 30-day survivors of ischaemic stroke treated with low dose aspirin, modified-release dipyridamole; the coformulation of low dose aspirin plus modified-release dipyridamole, or no antiplatelet therapy. RESULTS The model predicted that over 5 years the coformulation prevented 29 more strokes than aspirin alone per 1000 patients, at an additional cost of 1900 Pounds per stroke averted (1996 values). Over 5 years, each antiplatelet therapy was cost saving compared with no therapy. Results were sensitive to the cost of acute care, the cost of long term care of disabled stroke survivors, the effectiveness of therapy and the background risk of recurrent stroke. In sensitivity analyses, the cost effectiveness did not exceed 7000 Pounds per stroke averted or 11,000 Pounds per quality-adjusted life-year (QALY) gained, except when varying the effectiveness parameter. CONCLUSIONS Application of a decision-analytic model to the results of ESPS-2 indicated that first-line therapy with the coformulation of modified-release dipyridamole and low dose aspirin to patients with a previous ischaemic stroke is likely to generate significant health benefits at modest extra costs to health and social services. The extra costs of treatment are balanced by the savings in future costs of acute care and long term care of the disabled. Future economic evaluations in this area should pay particular attention to the cost perspective, the duration of analysis, the selection of trials from which effectiveness data are derived, and the impact of the pooling of outcome events with potentially different economic consequences.