In view of the lack of reliable economic information from most European countries, Sweden (population 8.5 million) is used here as an example of the cost-effectiveness of stroke care in Europe. In 1988, stroke patients in Sweden accounted for 3.25 million bed days in hospitals and nursing homes (382,000/million inhabitants). The total direct and indirect costs associated with stroke have been calculated for these patients, using 1991 prices. The total direct costs (made up mainly of medical care and social services costs) are calculated to be US$ 1,579 million (US$ 185.8 million/million inhabitants) and the total indirect costs (made up of sickness benefits, stroke-related early retirement and death before the age of 65 years) are calculated to be US$ 396 million (US$ 46.6 million/million inhabitants). The total calculated costs are therefore US$ 1,975 million (US$ 232.4 million/million inhabitants). Stroke patients are major consumers of medical care and social services even before the occurrence of stroke. Adjusting for this, the national incremental cost attributed to stroke is calculated to be US$ 1,017 million (US$ 120 million/million inhabitants). The average cost from first stroke to death is US$ 79,000 per patient (US$ 41,000 if adjusted for the high consumption of medical care and social services before stroke). The economic benefits of three types of preventive strategy have been calculated using the results of recent randomized trials applied to the Swedish medical setting. Anticoagulants given to eligible patients with atrial fibrillation might prevent 420 strokes/year, and give a net economic saving of US$ 1.5 million/year (US$ 0.2 million/million inhabitants), provided that the risk of intracranial bleeding is low (1.3%/year), as in well-controlled routine care in Sweden. Carotid surgery in patients with transient ischemic attack or minor stroke might prevent 255 strokes/year and save US$ 8 million/year (US$ 0.9 million/million inhabitants). Antiplatelet agents in patients with transient ischemic attack or minor stroke have the potential to prevent 180 strokes/year and save US$ 7.2 million/year (US$ 0.8 million/million inhabitants); these drugs now being used routinely, the US$ 7.2 million saving has already been made. Intensified efforts to identify and treat hypertension have a high potential for reducing the number of strokes (at least 1,000 prevented per year in Sweden), but the costs are considerable. Recent randomized trials have shown that antihypertensive therapy effectively reduces the risk of stroke and also of myocardial infarction in people above 70 years of age. In the 70- to 80-year age range, the net cost is approximately US$ 250/year gained. At 50 years, the corresponding economic saving approaches US$ 250/year gained. At 50 years, the corresponding economic saving approaches US$ 60,000/year. Nonpharmacological strategies to reduce blood pressure levels in the population and campaigns against smoking are probably cost-effective, but no data are yet available to support this proposition. Admitting stroke patients to specialized stroke units during the acute phase has no major influence on mortality. However, early intensive rehabilitation in such units reduces the need for long-term institutional care, as shown by two Scandinavian randomized trials. Ongoing analyses indicate that stroke units are also extremely beneficial from an economic perspective.