MicroRNA-27a (miR-27a) has been reported to be abnormally expressed in patients with cancer, and it could play potential roles as a diagnostic and prognostic biomarker of cancers. However, the diagnostic and prognostic role remains unclear. Hence, this meta-analysis, based on published data, was conducted to assess the utility of miR-27a as a diagnostic and prognostic marker in various cancers. To identify eligible studies, databases: Web of Science, PubMed, and CNKI were searched, with 868 literatures obtained, 16 of which were included in the Meta-analysis. The pooled results of studies conducted with serum/plasma showed that miR-27a was a valuable diagnostic biomarker in cancers (area under curve (AUC)= 0.91, sensitivity (SEN)= 0.84, specificity (SPE)= 0.85), with the diagnostic value slightly reduced in tumor tissue samples (AUC=0.83, SEN=0.78, SPE: 0.74). Additionally, the pooled results revealed that high expression of miR-27a predicted poor prognosis of cancer in serum/plasma (hazard ratio (HR) = 0.63, PHeterogeneity = 0.278, I2= 21.50%) but not in tumor tissue (HR = 0.98, PHeterogeneity =0.577, I2= 0.0). In brief, our results suggested that miR-27a in serum/plasma or tumor tissue could act as a diagnostic biomarker, and that miR-27a in serum/plasma could predict cancer patients’ survival.