One hundred and forty-seven patients had intracranial pressure monitoring by means of a subarachnoid screw as part of their routine neurosurgical care. Average length of monitoring was five days. Successful recordings were obtained in 92% of the patients. Improper technical placement caused the majority of failures. The overall infection rate associated with the monitor was 2.1%, with 0.7% CNS infection (one patient developed a brain abscess). We have since monitored an additional 500 patients and find the subarachnoid screw to be a reliable, low risk means of clinically recording intracranial pressure in neurosurgical patients.