A new method for the determination of hydraulic conductivity in individually perfused microvessels in vivo is described. A vessel is cannulated at both ends with glass micropipettes and the fluid filtration rate across the vessel wall measured from the velocities of red cells when the pressure in the micropipettes is balanced. Hydraulic conductivity measured using this double-cannulation method (2.6 (± 0.9) x 10^-7 cm s^-1 cmH₂O^-1) was not significantly different from that measured using a previously described technique in the same vessel (2.4 (± 0.9) x 10^-7 cm s^-1 cmH₂O^-1 using the Landis-Michel method). Shear stress on the vessel wall was controlled by changing the difference between the inflow and outflow pressures during periods of perfusion. The volume flow through the vessel, calculated from red cell velocity either in the vessel or in the pipette, was linearly proportional to this pressure difference. Higher flow rates could only be calculated from red cell velocities in the micropipette. There was no relationship between the imposed shear stress and intervening measurements of hydraulic conductivity (r = 0.029). This novel technique has advantages over the Landis-Michel method, which include the control of outflow resistance, the measurement of shear stress under conditions of controlled pressure, the elimination of compression damage to the vessel (since vessel occlusion is not necessary) and assessment of hydraulic conductivity over the same length of vessel throughout the experiment. The measurement of solute concentrations by indwelling micropipette electrodes and the collection of perfusate for analysis are other possibilities.