OBJECTIVE To investigate differences between in-vivo properties of a vascular bed in hypertensive patients and normotensive controls. METHODS Despite the controversy about the origin of essential hypertension and its accompanying vascular changes, it is generally assumed that the characteristic increase in peripheral resistance when hypertension progresses is caused by arteriolar constriction. Yet, there is little experimental evidence that this assumption generally holds in vivo. METHODS A non-invasive technique was used for studying properties of the complete vascular bed of an upper arm segment under an occluding cuff in 23 previously untreated hypertensive patients and their matched normotensive controls. The method used the segment's electrical impedance to assess the volumes of extravascular fluid and of arterial and venous blood under varying arterial transmural pressures. RESULTS Compared with that of matched normotensive controls, the compliance of the large arteries of the vascular bed was on average 50.9% lower (P < 0.001) in the hypertensive patients. The compliance of the complete arterial bed at the operating blood pressure level was also lower (40.0%, P < 0.01), but appeared to be significantly higher (45.9%, P < 0.05) at the normotensive blood pressure level. On the venous side, the patients had a higher blood volume (60.0%, P < 0.01) and an increased myogenic response (68.5%, P < 0.05). CONCLUSIONS The increase in vascular resistance in the hypertensive patients is due primarily to changes in the large and small vessels of the arterial bed. We found no evidence for a generally increased arteriolar constriction.