OBJECTIVE: The goal of this study was to analyze the use of thermal diffusion-cerebral blood flow (TD-CBF) monitoring as a continuous monitoring method for patients with head injuries in the intensive care unit. METHODS: The TD-CBF probe was placed on normal-appearing frontal or parietal cortex at the time of surgery to treat traumatic intracranial hematomas in 35 patients, and cortical cerebral blood flow (CBF) was monitored for up to 7 days after surgery. We compared TD-CBF values with global CBF values measured by the standard Kety-Schmidt technique, and we compared changes in TD-CBF with changes in jugular venous oxygen saturation observed during intracranial pressure elevations. RESULTS: The average value for the global CBF measurements was 50.5+/-0.9 ml/100 g/min and that for the TD-CBF measurements was 60.5+/-1.4 ml/100 g/min; the average difference was 9.3+/-1.2 ml/100 g/min. The overall slope of the regression between the global CBF and TD-CBF measurements (n = 206) was 0.636 (comparison of observed slope with a slope of 0, P < 0.001). The relationship between the TD-CBF and global CBF values during 546 episodes of increased intracranial pressure was examined by comparing the changes in TD-CBF with the changes in jugular venous oxygen saturation. When the change in TD-CBF was at least 10 ml/100 g/min during an intracranial pressure elevation, the TD-CBF change reflected the change in jugular venous oxygen saturation on 85% of the occasions. CONCLUSION: The TD-CBF method is very convenient because of the continuous and automatic nature of the measurements. Most of the time, a change in TD-CBF indicated a similar change in global CBF. However, the limitations of local measurements of CBF must be kept in mind during therapeutic decision-making.