The development of high-throughput technologies has yielded a large amount of data from molecular and epigenetic analysis that could be useful for identifying novel biomarkers of cancers. We analyzed Gene Expression Omnibus (GEO) DataSet micro–ribonucleic acid (miRNA) profiling datasets to identify miRNAs that could have value as diagnostic and prognostic biomarkers in hepatocellular carcinoma (HCC). We adopted several computing methods to identify the functional roles of these miRNAs. Ultimately, via integrated analysis of three GEO DataSets, three differential miRNAs were identified as valuable markers in HCC. Combining the results of receiver operating characteristic (ROC) analyses and Kaplan–Meier Plotter (KM) survival analyses, we identified hsa-let-7e as a novel potential biomarker for HCC diagnosis and prognosis. Then, we found via quantitative reverse-transcription polymerase chain reaction (RT-qPCR) that let-7e was upregulated in HCC tissues and that such upregulation was significantly associated with poor prognosis in HCC. The results of functional analysis indicated that upregulated let-7e promoted tumor cell growth and proliferation. Additionally, via mechanistic analysis, we found that let-7e could regulate mitochondrial apoptosis and autophagy to adjust and control cancer cell proliferation. Therefore, the integrated results of our bioinformatics analyses of both clinical and experimental data showed that let-7e was a novel biomarker for HCC diagnosis and prognosis and might be a new treatment target.