Thalidomide, an anti-inflammatory and immunomodulatory agent, has a potential role in cases with central nervous system tuberculosis (CNS-TB) with paradoxical reactions. Although several articles have described the use of thalidomide in CNS-TB, no systematic review has been performed in this regard. Different electronic databases were searched for articles describing the use of thalidomide in patients with CNS-TB. For determining pooled estimates in the quantitative review, studies with a minimum sample size of 5 were only considered, whereas for qualitative synthesis even single case reports were included. Fixed or random effect models were used suitably depending on the degree of heterogeneity. Fourteen articles describing a total of 107 patients (98 children and 9 adults) were selected from 156 records. A favorable clinical response was observed in 89% of patients with CNS-TB who had paradoxical reactions refractory to corticosteroids. Majority of the studies used a dose of 2–6 mg/kg/day and around 24% suffered from at least one adverse effect, with a mortality of 5%. Predominant adverse effects were rash (9.5%), neuropathy (6%), and elevated liver transaminases (9.5%). Only one placebo-controlled trial has been performed till now, which showed that high-dose thalidomide has numerous adverse effects, without any clinically significant improvement as compared with placebo. While in HIV-positive patients with TB-immune reconstitution inflammatory syndrome thalidomide was helpful in around 82% of cases. Low-dose thalidomide is helpful in patients with CNS-TB who had a paradoxical reaction and unresponsive to corticosteroids. Large, randomized trials are needed to provide more concrete information regarding the safety and efficacy of thalidomide.