Accurate quantification of miRNA expression level is essential to the study of its biology, and many cutting-edge technologies have been developed to accommodate this need. Yet most of them were designed primarily for the "regular" RNAs such as animal miRNAs and may overlook the fact that plant miRNAs and many other small noncoding RNAs are 2'-O-methylated at the 3' end nucleotide. According to our experimental data and previous reports, this structural variation is detrimental to the effectiveness of the commonly used enzymatic labeling methods, leading to strongly biased results (~24-fold difference). Herein, we demonstrate that our Stacking-Hybridized Universal Tag (SHUT) microarray assay is well suited for unbiased profiling of both normal and methylated small RNA species. The detected signals of small RNAs with 2'-hydroxyl and 2'-O-methyl 3' ends are highly consistent (no significant difference at α = 0.01 level). For specificity, the presented method edges over others by its unique ability to discriminate single-base difference at or near the 5' end. Notably, as compared to many delicate techniques, this enzyme-free and label-free approach requires much less reagent and manipulation, benefiting the SHUT-based applications with more efficient workflow and highly reproducible results.