We investigated the potential role of serum procalcitonin in differentiating tuberculosis meningitis from bacterial and viral meningitis, and in predicting the prognosis of tuberculosis meningitis.This was a retrospective study of 26 patients with tuberculosis meningitis. In addition, 70 patients with bacterial meningitis and 49 patients with viral meningitis were included as the disease control groups for comparison. The serum procalcitonin level was measured in all patients at admission. Differences in demographic and laboratory data, including the procalcitonin level, were analyzed among the three groups. In addition, we analyzed the predictive factors for a prognosis of tuberculosis meningitis using the Glasgow Coma Scale (GCS) at discharge, and the correlation between the level of procalcitonin and the GCS score at discharge.Multiple logistic regression analysis showed that a low level of procalcitonin (≤1.27 ng/mL) independently distinguished tuberculosis meningitis from bacterial meningitis. The sensitivity and specificity for distinguishing tuberculosis meningitis from bacterial meningitis were 96.2% and 62.9%, respectively. However, the level of procalcitonin in patients with tuberculosis meningitis did not differ significantly from that in patients with viral meningitis. In patients with tuberculosis meningitis, a high level of procalcitonin (>0.4 ng/mL) was a predictor of a poor prognosis, and the level of procalcitonin was negatively correlated with the GCS score at discharge (r=-0.437, p=0.026).We found that serum procalcitonin is a useful marker for differentiating tuberculosis meningitis from bacterial meningitis and is also valuable for predicting the prognosis of tuberculosis meningitis.