BACKGROUND The evidence-based approach to medical care involves the explicit use of evidence on the magnitude of the effects of interventions to inform diagnostic and treatment decisions. This article critiques current mainstream guidelines on the management of hypertension in the elderly (aged 60 years and over) and presents an alternative evidence-based approach. METHODS Three major national and international guidelines for the management of hypertension from the United Kingdom (UK), the United States (US) and from a joint World Health Organisation/International Society of Hypertension (WHO/ISH) Working Party were appraised and the evidence on which they were based was reviewed. The relevant evidence was also assessed to determine the likely magnitude of risks and benefits of anti-hypertensive treatment in older people and an alternative approach to making treatment decisions, based on the New Zealand guidelines for the management of hypertension, is described. RESULTS Hypertension management guidelines from the UK, US and WHO/ISH made similar recommendations about which elderly patients should be treated, although there were some ambiguities in their advice. Treatment recommendations were based primarily on blood pressure levels which were set at about 160 mm Hg systolic and/or 90 mm Hg diastolic. The threshold levels were based mainly on the cut-off blood pressure levels used in randomised trials of anti-hypertensive drug treatment, rather than the estimated magnitude of treatment benefit. Each of the guidelines acknowledged the important effect of associated cardiovascular disease (CVD) risk factors on the likely benefits of treatment, but did not expand on the magnitude of this effect. No patient-specific estimates of the likely absolute benefits of treatment were provided in any of the guidelines. In contrast the New Zealand guidelines for the management of hypertension recommend the use of explicit estimates of absolute CVD risks and benefits to inform treatment decisions. They were designed to provide practitioners with estimates of the likely absolute risk of CVD in patients with different risk factor profiles and with estimates of the absolute benefits of treatment. The New Zealand guidelines recommend that drug treatment be considered in patients with a 5-year risk of CVD of about 10-15% or more; approximately 25 patients with a 10-15% risk would require treatment for 5 years to prevent one CVD event. As elderly patients are generally at higher absolute CVD risk than younger people, the New Zealand recommendation give priority to the treatment of older patients. In order to take account of differences in life expectancy and the medical costs of caring for elderly people, absolute risk-based guidelines can be improved by incorporating potential years of life gained from treatment and the cost-effectiveness of treatment expressed as $/quality adjusted life years gained. Preliminary analyses indicate that the cost-effectiveness of treatment is generally greatest in patients in their 60s and early 70s. Treatment in younger people is not usually very cost-effective because of their low absolute risk of CVD and the cost-effectiveness of treatment in people over about 75 years declines because of the increasing cost of non-CVD morbidity. CONCLUSIONS The explicit assessment of absolute CVD risks and likely treatment benefits in patients with hypertension can usefully inform treatment decisions and provide a more rational basis for initiating therapy than blood pressure levels alone. This approach highlights the generally greater CVD risk and potential treatment benefits in older compared with younger hypertensive patients. The absolute risk-based approach can be further enhanced by providing decision makers with patient-specific data on the potential life years gained from treatment and its cost-effectiveness. (ABSTRACT TRUNCATED)