Non-invasive aortic compliance measurements have been used previously to assess the distensibility of the aorta in several pathological conditions associated with increased cardiovascular risk. We set out to establish whether aortic compliance is abnormal in patients with stroke. 2. Pulse wave velocity measurements of thoraco-abdominal aortic compliance were made in 20 stroke patients and 25 age- and sex-matched hospitalized, non-stroke control subjects putatively free of cardiovascular disease. Since compliance varies with non-chronic changes in blood pressure, a blood pressure corrected index of aortic distensibility, Cp, was calculated. 3. Aortic compliance was significantly reduced in patients with stroke compared with non-stroke control subjects (0.46 +/- 0.27 versus 0.86 +/- 0.34%/10 mmHg, P < 0.0002), corresponding with higher values for pulse wave velocity. Stroke patients also had significantly higher systolic and diastolic blood pressures (P < 0.02 and P < 0.002 respectively) and total cholesterol levels (P < 0.004) than the control subjects. Calculation of Cp did not alter the observation of stiffer aortas in the stroke cohort (P < 0.0007). 4. In both stroke patient and control cohorts, as expected, inverse trends were observed between aortic compliance and blood pressure. Also as expected, in the control group Cp values did not show a relationship with blood pressure (r = 0.02, P = 0.092, not significant). However, in the stroke cohort a marked dependence of Cp on blood pressure was observed (r = -0.48, P = 0.03). 5. Transoesophageal echocardiographic studies have recently identified advanced atherosclerosis in the ascending aorta as a possible source of cerebral emboli and an independent risk factor for ischaemic stroke.(ABSTRACT TRUNCATED AT 250 WORDS)