An increase in the number of cases of gram-negative ventriculomeningitis in patients followed with intracranial pressure monitors when compared with patients with craniotomy alone was revealed by routine surveillance data. A study was undertaken at four area institutions to describe the infections, risk factors, and management. Two hundred fifty-five patients with diagnoses of intracerebral hemorrhage (n = 86), closed trauma (n = 66), open trauma (n = 21), tumor (n = 66), and miscellaneous other conditions were compared with their nonmonitored counterparts for type of intracranial pressure monitor used, use of drains, prophylactic antibiotics, and steroids, and remote presence of infection. The presence of intracranial pressure monitor with craniotomy was associated with an 11 percent infection rate whereas craniotomy alone demonstrated a 6 percent rate. Of the intracranial pressure monitors used, the subarachnoid screw was associated with the lowest infection rate (7.5 percent) followed by the subdural cup catheter (14.9 percent) and the ventriculostomy catheter (21.9 percent). Regardless of the monitor used, infection was twice as likely to develop in patients with open trauma or hemorrhage. The use of bacitracin flush solutions for maintenance of lumen patency was more often associated with infections. Use of prophylactic antibiotics did not significantly influence outcome.