In a series of studies (cf. Folkow et al. 1973) the design, and reactivity, of the resistance vessels in hypertension has been hemodynamically analyzed, comparing constant flow perfusion of, preferentially, the hindquarter vascular beds of hypertensive (SHR) and paired normotensive control rats (NCR). Measurements of the resistance at maximal dilatation, the 'threshold sensitivity' to noradrenaline (NA), the entire dose-response curve to NA and its steepness, and the maximal pressor response to supramaximal concentrations of constrictor agents, revealed important differences in resistance vessel design. Thus, in both SHR and rats with renovascular hypertension ( R H R ) , resistance vessel design is rapidly changed towards a luminal narrowing associated with a hypertrophic wall increase. This type of 'structural autoregulation', of the resistance vessels, triggered by functional pressor influences, appears to be of great importance in hypertension since resistance is then raised also when smooth muscle tone is normal.To get more insight into hypertensive vascular beds, methods are needed allowing separate quantitative estimations of e.g. precapillary resistance vessels, capillary exchange vessels, postcapillary resistance and capacitance vessels, which are specialized both in design and function. For such purposes hindquarters of rats were isolated by thorough ligations of all vascular connections except the aorta and caval vein.Arterial inflow pressure (4\) was measured in the tail artery and venous outflow pressure (Pv) in a caval branch. The caval vein was connected to a wide outflow tube whereby P,could be set at desired levels. The preparation was connected to a constant flow perfusion pump delivering a body-warm plasma substitute (oxygenated Tyrode solution with 4 % Ficoll, m.w. about 80,000, AB Pharmacia, Sweden), a t known flows (Q) and levels of P.% and Pv. The hindquarters were connected to a force displacement transducer allowing continuous, precise recordings of weight shifts of the preparation. From these shifts, changes in intravascular fluid content and in transcapillary fluid transfer could be separately deduced (for details see e.g. Mellander 1960). I n this way PA,, Pv and Q could be set a t desired levels and 1 Inst. of Physiology, Czechoslovak Academy of Sciences, Praha; guest research scientist of the Royal Academy of Sciences, Sweden.