OBJECTIVE: To compare the characteristics of patients with cerebrovascular disease in primary care with those of the participants in the PROGRESS trial, on which national guidelines for blood pressure lowering are based. DESIGN: Population based cross sectional survey of patients with confirmed stroke or transient ischaemic attack. SETTING: Seven general practices in south Birmingham, England. PARTICIPANTS: All patients with a validated history of stroke (n = 413) or transient ischaemic attack (n = 107). MAIN OUTCOME MEASURES: PATIENT CHARACTERISTICS: age, sex, time since last cerebrovascular event, blood pressure, and whether receiving antihypertensive treatment. RESULTS: Patients were 12 years older than the participants in PROGRESS and twice as likely to be women. The median time that had elapsed since their cerebrovascular event was two and a half years, compared with eight months in PROGRESS. The systolic blood pressure of 315 (61%) patients was over 140 mm Hg, and for 399 (77%) it was over 130 mm Hg. One hundred and forty seven (28%) patients were receiving a thiazide diuretic, and 136 (26%) were receiving an angiotensin converting enzyme inhibitor. CONCLUSIONS: Important differences exist between the PROGRESS trial participants and a typical primary care stroke population, which undermine the applicability of the trial's findings. Research in appropriate populations is urgently needed before the international guidelines are implemented in primary care.