Neuroinflammation has been implicated in the neurobiological pathways of schizophrenia. This meta-analysis evaluated the efficacy and tolerability of adjunctive celecoxib as a noncompetitive anti-inflammation drug in treating schizophrenia. Data were searched, extracted, checked and entered into the RevMan (version 5.3) software by two independent investigators. Standardized/weighted mean differences (SMDs/WMDs), risk ratio (RR) and their 95% confidence intervals (CIs) were calculated, as appropriate. Included were 8 randomized controlled trials (RCTs) with a total of 626 patients with schizophrenia including 316 (50.5%) treated on celecoxib (400 mg/day) and 310 (49.5%) on placebo over 8.3 ± 2.3 weeks of treatment. Adjunctive celecoxib outperformed placebo with respect to total psychopathology [3 RCTs, n = 180; SMD: −0.47; 95%CI: −0.81, −0.14; P = 0.005; I2 = 18%; 'moderate quality'], symptoms positive [3 RCTs, n = 180; SMD: −0.50; 95%CI: −0.79, −0.20; P = 0.001; I2 = 0%; 'moderate quality'], negative symptoms [3 RCTs, n = 180; SMD: −0.32; 95%CI: −0.66, 0.02; P = 0.06; I2 = 22%; 'moderate quality'], and general psychopathology scores [3 RCTs, n = 180; SMD: −0.35; 95%CI: −0.65, −0.06; P = 0.02; I2 = 0%; 'moderate quality'] in first-episode, but not chronic patients. Additionally, superiority of celecoxib for the Positive and Negative Syndrome Scale (PANSS) total scores was moderated by higher PANSS positive scores and lower PANSS negative scores at baseline. All-cause discontinuation [RR: 1.02; (95%CI: 0.56, 1.86); P = 0.94; I2 = 0%] and adverse drug reactions were similar between the two groups. Adjunctive celecoxib appears to be an efficacious and safe treatment in improving psychotic symptoms, particularly in first-episode schizophrenia. Review registration CRD42016054233 (http://www.crd.york.ac.uk/prospero/).