Multiple system atrophy (MSA) is an adult onset, fatal neurodegenerative disease. However, no reliable biomarker is currently available to guide clinical diagnosis and help to determine the prognosis. Thus, a comprehensive meta-analysis is warranted to determine effective biomarkers for MSA and provide useful guidance for clinical diagnosis.A comprehensive literature search was made of the PubMed, Embase, Cochrane and Web of Science databases for relevant clinical trial articles for 1984-2019. Two review authors examined the full-text records, respectively, and determined which studies met the inclusion criteria. We estimated the mean difference, standard deviation and 95% confidence intervals.A total of 28 studies and 11 biomarkers were included in our analysis. Several biomarkers were found to be useful to distinguish MSA patients from healthy controls, including the reduction of phosphorylated tau, -synuclein (-syn), 42-amino-acid form of A and total tau (t-tau), the elevation of neurofilament light-chain protein (NFL) in cerebrospinal fluid, the elevation of uric acid and reduction of homocysteine and coenzyme Q10 in plasma. Importantly, -syn, NFL and t-tau could be used to distinguish MSA from Parkinson's disease (PD), indicating that these three biomarkers could be useful biomarkers in MSA diagnosis.The findings of our meta-analysis demonstrated diagnostic biomarkers for MSA. Moreover, three biomarkers could be used in differential diagnosis of MSA and PD. The results could be helpful for the early diagnosis of MSA and the accuracy of MSA diagnosis.