Blood pressure and components of blood viscosity were measured in 49 normal subjects and in 49 untreated patients with essential hypertension. Blood viscosity values measured at six different shear rates were significantly correlated with blood pressure (r = 0.432 to 0.505, p less than 0.001). Blood viscosity was higher in hypertensive patients. This was due to both higher plasma viscosity (1.29 +/- 0.08 standard deviation versus 1.24 +/- 0.05 centipoise (cPs), p less than 0.001) and increased hematocrit values (44.4 +/- 4 percent versus 41.5 +/- 3 percent, p less than 0.005). When blood viscosity was evaluated in subgroups of normal and hypertensive subjects with matched hematocrit values, it remained higher in the hypertensive patients, and the relationship between blood pressure and viscosity was still significant. Regardless of the hematocrit value, fibrinogen levels were elevated in hypertensive patients (p less than 0.006) and, in association with the increased globulin concentration, fibrinogen was largely responsible for the increased plasma viscosity in hypertensive patients. Since the viscosity of defibrinated blood was similar in normal and hypertensive subjects with matched hematocrit values, the elevated fibrinogen level also affected whole blood viscosity. Defibrinated blood viscosity and arterial pressures were not correlated. These studies demonstrate a direct correlation between blood pressure and blood viscosity among normotensive and hypertensive subjects. This relationship is, in part, due to the rheologic effects of an elevated fibrinogen level and to an increased hematocrit value. The basis for hyperfibrinogenemia in hypertensive patients is unclear.