Objective: As a non-competitive N-methyl-D-aspartate receptor antagonist, memantine has been used to treat major mental disorders including schizophrenia, bipolar disorder, and major depressive disorder (MDD). This meta-analysis systematically investigated the effectiveness and tolerability of adjunctive memantine for patients with schizophrenia, bipolar disorder, and MDD. Methods: Only randomized controlled trials (RCTs) were identified and included in the study. Data of the three disorders were separately synthesized using the RevMan 5.3 software. Results: Fifteen RCTs (n = 988) examining memantine (5–20 mg/day) as an adjunct treatment for schizophrenia (9 trials with 512 patients), bipolar disorder (3 trials with 319 patients), and MDD (3 trials with 157 patients) were analyzed. Memantine outperformed the comparator regarding total psychopathology with a standardized mean difference (SMD) of −0.56 [95% confidence interval (CI): −1.01, −0.11; I2 = 76%, P = 0.01] and negative symptoms with an SMD of −0.71 (95% CI: −1.09, −0.33; I2 = 74%, P = 0.0003) in schizophrenia, but no significant effects were found with regard to positive symptoms and general psychopathology in schizophrenia, or depressive and manic symptoms in bipolar disorder or depressive symptoms in MDD. Memantine outperformed the comparator in improving cognitive performance in schizophrenia with an SMD of 1.07 (95% CI: 0.53, 1.61; P < 0.0001, I2 = 29%). No group differences were found in the rates of adverse drug reactions and discontinuation due to any reason in the three major mental disorders. Conclusions: Memantine as an adjunct treatment appears to have significant efficacy in improving negative symptoms in schizophrenia. The efficacy and safety of adjunctive memantine for bipolar disorder or MDD needs to be further examined. Review registration: PROSPERO: 42018099045.