MicroRNAs (miRNAs) are a class of short, endogenous, non-coding small RNAs. The regulation function accompanies with the cell growth, differentiation, proliferation and apoptosis. The miRNAs expression is closely related with the development of cancer and has been considered as a potential biomarker. The analysis of the miRNAs expression profile is a critical part in the studying of miRNAs. Usually, miRNAs in biological samples require fraction, labeling and purification before applied to most assays. This is the most time-consuming, labor-intensive and the highest cost in the assay process. The assay results would be affected for the initial ratio of the sample target miRNAs which may have been changed due to the application of enzymes and complicated steps in this preprocess. The main purpose of this work is to develop a new microRNA microarray platform that free of the sample labeling. It is named the stacking hybridization-based universal tag (SHUT) assay for it takes advantage of stacking hybridization interaction and involves a universal tag (UT)——an 8mers oligonucleotide labeled previously. This article focuses on the optimization of the experimental procedures in the SHUT assay and the evaluation of its properties, such as the sensitivity and the specificity. The results show that the microarray has the advantage of high sensitivity, that even 2 fmol/L input miRNAs can be detected. The miRNAs microarray has perfect selectivity simultaneously. It can distinguish the target miRNAs from other family members with only one base mismatched. In particularly, the pri-miRNA and pre-miRNA can be easily discriminated from mature-miRNAs, which enables as little as 100 ng total RNAs to be analyzed directly and rapidly. All these results have shown that the application of this new microarray platform approach is a rapid and ideal platform for the detection as well as analysis of miRNAs and other small nucleic acid molecules.