BACKGROUND: Prostate cancer (PC) is the second most common malignant tumor in males and a leading cause of cancer-related morbidity and mortality among men. Early detection is essential for improving outcomes. Although prostate-specific antigen (PSA) is widely used for PC screening, it suffers from high false positive and false negative rates. Our study aimed to identify a panel of serum mature microRNAs (miRNAs) for PC diagnosis. METHODS: We conducted a PubMed search to identify candidate mature miRNAs associated with PC. We then used quantitative reverse transcription-polymerase chain reaction (RT-qPCR) to assess the expression profiles of these mature miRNAs in serum samples from 112 PC patients and 112 healthy controls (HCs). We selected mature miRNAs with favorable diagnostic potential by analyzing receiver operating characteristic (ROC) curves and calculating the area under the curve (AUC). Subsequently, we developed a highly diagnostically efficient panel of three mature miRNAs using stepwise logistic regression based on their expression levels. RESULTS: We identified three mature miRNAs (hsa-miR-143-5p, hsa-miR-23b-3p, and hsa-miR-148b-3p) with significant diagnostic value, constructing a panel with an AUC of 0.891, sensitivity of 84.15%, and specificity of 80.49%. Bioinformatics analysis also revealed LDB3 and RBMS3 as potential therapeutic targets for PC. CONCLUSIONS: Our study introduces a novel diagnostic approach by identifying a panel of three mature miRNAs (hsa-miR-143-5p, hsa-miR-23b-3p, and hsa-miR-148b-3p) as novel and non-intrusive biomarkers for PC.