Wall shear stress (WSS), the drag of the flowing blood exerted on endothelial cells, is an important determinant of endothelial cell function and gene expression. WSS can be estimated from wall shear rate (WSR) and local blood viscosity, WSR being defined as the radial derivative of blood flow velocity at the wall. In large arteries in man, WSR is derived from velocity profiles, non-invasively recorded by means of ultrasound or Magnetic Resonance Imaging (MRI). In our studies we use ultrasound, because of its better spatial and especially temporal resolution than of MRI. The velocity profiles are recorded with a two-dimensional ultrasound imaging device combined with a dedicated acquisition and processing system as developed in our institute. WSS is estimated from the derived WSR and whole blood viscosity, because the thin plasma layer can be ignored relative to the size of the ultrasound sample volume (sample length 300 µm; 50% overlapping). In arterioles WSS is measured directly or estimated from WSR and plasma viscosity, WSR being derived from velocity profiles recorded with labeled blood platelets or nanometer particles as velocity tracers. Originally, the displacement and the radial position of the velocity tracers were determined by hand, a time consuming procedure. Recently a computerized two-dimensional particle tracking technique has been developed to determine radial position and displacement of the particles, automatically providing velocity profiles. In arterioles plasma viscosity can be used to calculate WSS, because the velocity tracers come as close to the wall as 0.2-0.5 µm. The in vivo measurements have shown that the theoretical assumptions regarding WSS in the arterial system and its calculation are far from valid. In both arteries and arterioles, velocity profiles are flattened rather than fully developed parabolas. This implies that WSR has to be derived from recorded velocity profiles. Assuming a parabolic velocity profile will on the average underestimate derived WSR by a factor of 2-3. In humans mean WSS varies along the arterial tree and is higher in the common carotid artery (1.1-1.3 Pa; 1 Pa=10 dyn●cm^-2) than in the brachial artery (0.4- 0.5 Pa) and the common (0.3-0.4) and superficial (0.5 Pa) femoral arteries. Only in the common carotid artery mean WSS is close to the theoretically predicted value of 1.5 Pa. The lower mean WSS in conduit arteries can be explained by the high peripheral resistance in these arteries, reducing mean volume flow and inducing reflections. Dilation of the femoral artery vascular bed results in mean WSS values in this artery not significantly different from those in the common carotid artery. This observation indicates that at rest mean WSS is largely determined locally. Although small, the difference in mean WSS between the common and the superficial femoral artery is significant, the former artery seeing reflections from both the deep and the superficial artery, while the latter one only sees reflections from its own vascular bed. Also in the carotid artery bifurcation differences in mean WSS have to be appreciated. It is of interest to note that in both the femoral and the carotid artery bifurcation the differences in mean WSS are associated with local differences in intima-media thickness (IMT): the lower mean WSS is, the larger IMT will be. Also in animals mean WSS is not constant along the arterial tree. In arterioles mean WSS varies between 2 and 10 Pa and is dependent on the site of measurement in the arteriolar network. Across species mean WSS in a particular artery decreases linearly with increasing body mass on a log-log scale, in the infra-renal aorta from on the average 8.8 Pa in mice to 7.0 Pa in rats and 0.5 Pa in humans (flow velocities being similar). A similar pattern can be found in the carotid artery, varying on the average from 7.0 Pa in mice to 4.7 Pa in rats and 1.2 Pa in man. The observation that mean WSS is far from constant along the arterial tree indicates that Murray’s cube law on flow-diameter relations cannot be applied to the whole arterial system. At the present state of the art it can be concluded that the exponent of the power law varies from 2 in large branches of the aortic arch to 2.55 in coronary arteries and 3 in arterioles. The in vivo findings also imply that in in vitro investigations no average calculated shear stress value can be taken to study gene expression by endothelial cells derived from different vascular areas or from the same artery in different species. The cells have to be studied under the shear stress conditions they are exposed to in real life. Sensing and transduction of shear stress is likely to be in part mediated by the endothelial glycocalyx, because pretreatment of endothelial cells with hyaluronidase, leading to substantial reduction of glycocalyx dimensions, attenuates shear stress induced release of nitric oxide and shape changes of these cells. Therefore, modulation of shear stress sensing and transduction by altered glycocalyx properties, for example, in atherogenesis, should be considered.