BACKGROUND: Polysaccharide pneumococcal vaccination for older adults is being introduced in developed country settings. Evidence of protection by this vaccine against pneumococcal pneumonia, or confirmation that illness and death from bacteraemia are prevented, is currently limited. Decisions are often made based on partial information. We examined the policy implications by exploring the potential economic benefit to society and the health sector of pneumococcal vaccination in older adults. METHODS: A model to estimate the potential cost savings and cost-effectiveness of a polysaccharide pneumococcal vaccine programme was based on costs collected from patients, the literature, and routine health-services data. The effect of a pneumococcal vaccine (compared with no vaccination) was examined in a hypothetical cohort aged over 64 years. The duration of protection was assumed to be 10 years, with or without a booster at 5 years. RESULTS: If it were effective against morbidity from pneumococcal pneumonia, the main burden from pneumococcal disease, the vaccine could be cost-neutral to society or the health sector at low efficacy (28 and 37.5%, respectively, without boosting and with 70% coverage). If it were effective against morbidity from bacteraemia only, the vaccine's efficacy would need to be 75 and 89%, respectively. If protection against both morbidity and mortality from pneumococcal bacteraemia was 50%, the net cost to society would be 2500 pounds per year of life saved ( 3365 pounds from the health-sector perspective). Results were sensitive to incidence, case-fatality rates, and costs of illness. CONCLUSIONS: A vaccine with moderate efficacy against bacteraemic illness and death would be cost-effective. If it also protected against pneumonia, it would be cost-effective even if its efficacy were low.