The viscosity of human blood relative to distilled water observed in a capillary viscometer decreases with the length (in the range of 1.65– 10.5 cm) and increases with the reduction of the radius (in the range of .0106–.0217 cm) of the capillaries. No such changes are present when the flow rate of oils instead of water is used for the calculation of relative viscosity. Since homogeneous Newtonian oils exhibit a viscosity pattern similar to that of human blood, the existence of an "anomalous viscosity" of human blood described by Fahraeus and Lindqvist was not confirmed.