To determine blood viscosity in adult comatose patients treated with mild therapeutic hypothermia after cardiac arrest and to assess the relation between blood viscosity, cerebral blood flow, and cerebral oxygen extraction. Observational study. Tertiary care university hospital. Ten comatose patients with return of spontaneous circulation after out-of-hospital cardiac arrest. Treatment with mild therapeutic hypothermia for 24 hours followed by passive rewarming to normothermia. Median viscosity at shear rate 50/s was 5.27 mPa · s (4.29-5.91 mPa · s) at admission; it remained relatively stable during the first 12 hours and decreased significantly to 3.00 mPa · s (2.72-3.58 mPa · s) at 72 hours (p < 0.001). Median mean flow velocity in the middle cerebral artery was low (27.0 cm/s 23.8-30.5 cm/s) at admission and significantly increased to 63.0 cm/s (51.0-80.0 cm/s) at 72 hours. Median jugular bulb saturation at the start of the study was 61.5% (55.5-75.3%) and significantly increased to 73.0% (69.0-81.0%) at 72 hours. Median hematocrit was 0.41 L/L (0.36-0.44 L/L) at admission and subsequently decreased significantly to 0.32 L/L (0.27-0.35 L/L) at 72 hours. Median C-reactive protein concentration was low at admission (2.5 mg/L 2.5-6.5 mg/L) and increased to 101 mg/L (65-113.3 mg/L) in the following hours. Median fibrinogen concentration was increased at admission 2,795 mg/L (2,503-3,565 mg/L) and subsequently further increased to 6,195 mg/L (5,843-7,368 mg/L) at 72 hours. There was a significant negative association between blood viscosity and the mean flow velocity in the middle cerebral artery (p = 0.0008). Changes in blood viscosity in vivo are associated with changes in flow velocity in the middle cerebral artery. High viscosity early after cardiac arrest may reduce cerebral blood flow and may contribute to secondary brain injury. Further studies are needed to determine the optimal viscosity during the different stages of the postcardiac arrest syndrome.