To uncover the role of microRNAs in the occurrence and development of uveal melanoma (UM), we used R language packages in this study to analyze the correlations between the expression of microRNA isoforms, their target genes, and the clinical data for UM patients retrieved from The Cancer Genome Atlas (TCGA). We used Weighted Correlation Network Analysis (WGCNA) to divide the expression profiles of different microRNAs into 10 modules, among which blue and yellow modules were associated with UM survival. Hsa-miR-513a-5p, miR-506-3p, miR-508-3p, miR-140-3p, and miR-103a-2-5p were further identified as the top 5 node microRNAs based on the risk scores in both modules using least absolute shrinkage and selection operator (LASSO) Cox regression analysis. After combining these 5 microRNAs into an integrated risk signature, the prognostic performance of the risk signature was evaluated by area under the receiver operating characteristic (AUROC) curve, and their association with UM clinical characteristics was further analyzed using multiple Cox regression. Our results showed that this risk signature was sensitivity and specificity, and could serve as an independent prognostic factor. In addition, Spearman correlation analysis showed that expression of almost all target mRNAs were significantly positively or negatively correlated with the associated microRNAs. The gene ontology (GO), pathways, and disease enrichment analyses also showed that these 5 microRNAs were closely related to the incidence and progression of tumor, indicating their potential for predicting the outcome of UM. Abbreviations: AUROC = receiver operating characteristic, FPKM = fragments per kilobase of exon model per million reads mapped, GO = curve gene ontology, HR = Hazard ratio, KEGG = Kyoto Encyclopedia of Genes and Genomes, LASSO = least absolute shrinkage and selection operator, MCC = Matthews correlation coefficient, OS = overall survival, TCGA = The Cancer Genome Atlas, UM = Uveal melanoma, WGCNA = Weighted Correlation Network Analysis.