The differentiation of bacterial from aseptic meningitis in postoperative neurosurgical patients has traditionally been based on the clinical setting, a recent history of steroid administration, and cerebrospinal fluid (CSF) studies, including the total and differential leukocyte counts, Gram stain, glucose, and total protein. Recent reports questioning both the validity of a relative CSF lymphocytosis in excluding bacterial meningitis and the usefulness of standard CSF testing prompted the authors to reevaluate these standard criteria. The type of operation, the presence of a foreign body, use of steroids, postoperative day on which symptoms developed, altered mental status, neck stiffness, headache, and nausea were not helpful in the differential diagnosis. High fever, new neurological deficits, an active CSF leak, and elevated leukocyte counts in the CSF and peripheral blood favored a bacterial etiology. The CSF glucose level and the differential leukocyte count were less helpful. No criterion or combination of criteria was sensitive and specific enough to reliably differentiate aseptic from bacterial meningitis in the majority of patients. The possibility of improving diagnostic accuracy with newer tests, such as CSF lactate, ferritin, total amino acids, C-reactive protein, and amyloid-A, should be assessed.