Phosphorus magnetic resonance spectroscopy (P-31 MRS) allows in vivo quantification of phosphorus metabolites that are considered to be related to membrane turnover and energy metabolism. In schizophrenia (SZ), P-31 MRS studies found several abnormalities in different brain regions suggesting that alterations in these pathways may be contributing to the pathophysiology. In this paper, we systematically reviewed the P-31 MRS studies in SZ published to date by taking patient characteristics, medication status and brain regions into account. Publications written in English were searched on http://www.ncbi.nlm.nih.gov/pubmed/, by using the keywords 'phosphomonoester', 'phosphodiester', 'ATP', 'phosphocreatine', 'phosphocholine', 'phosphoethanolamineVglycerophosphocholine', 'glycerophosphoethanolamine', 'pH', 'schizophrenia', and 'MRS'. Studies that measured P-31 metabolites in SZ patients were included. This search identified 52 studies. Reduced PME and elevated PDE reported in earlier studies were not replicated in several subsequent studies. One relatively consistent pattern was a decrease in POE in chronic patients in the subcortical structures. There were no consistent patterns for the comparison of energy related phosphorus metabolites between patients and controls. Also, no consistent pattern emerged in studies seeking relationship between P-31 metabolites and antipsychotic use and other clinical variables. Despite emerging patterns, methodological heterogeneities and shortcomings in this literature likely obscure consistent patterns among studies. We conclude with recommendations to improve study designs and P-31 MRS methods in future studies. We also stress the significance of probing into the dynamic changes in energy metabolism, as this approach reveals abnormalities that are not visible to steady-state measurements. (C) 2015 Elsevier Ltd. All rights reserved.