Although hypofunction of N-methyl-D-aspartate (NMDA) receptor-mediated neurotransmission is proposed to play an important role in the pathophysiology of schizophrenia, results of the clinical trials of small molecules that enhance the NMDA function are inconsistent. A meta-analysis of all the double-blind, placebo-controlled studies in patients with schizophrenia was performed to examine their efficacy on different symptom domains, the dose-response, the effects of concomitant antipsychotics, and their side effects. About eight hundred subjects from 26 studies were included in current meta-analysis. Overall, the NMDA-enhancing molecules are effective in most schizophrenic symptom domains with the effect size (ES) of total psychopathology of 0.40 (p < 1x10-4). The ES of clinical efficacy of the symptom domains were in the order of depressive (0.40, p=3x10-4), negative (0.38, p < 1x10-4), cognitive (0.28, p=2x10-3), positive symptom (0.26, p=0.0006), and general psychopathology (0.26, p=0.006). Glycine, D-serine, and sarcosine treatments significantly improved multiple symptom domains, whereas D-cycloserine did not improve any symptom domain. Moderator analysis revealed that glycine, D-serine and sarcosine are better than D-cycloserine in improving the overall psychopathology. Patients receiving risperidone or olanzapine, but not clozapine, improved. No significant side effect or safety concern was noted. In addition to testing more lead compounds, long-term trials are required to determine their functional improvement capacity. Other drug targets that may enhance NMDA neurotransmission other than the molecules tested so far need to be explored.