Hereditary gingival fibromatosis (HGF) is a highly genetically heterogeneous disease, and current therapeutic method is limited to surgical resection with a high recurrence rate. MicroRNAs (miRNAs) are able to fine-tune large-scale target genes. Here we established a simple but effective computational strategy based on available miRNA target prediction algorithms to pinpoint the most potent miRNA that could negatively regulate a group of functional genes. Based on this rationale, miR-335-3p was top ranked by putatively targeting 85 verified profibrotic genes and 79 upregulated genes in HGF patients. Experimentally, downregulation of miR-355-3p was demonstrated in HGF-derived gingival fibroblasts as well as in transforming growth factor β–stimulated normal human gingival fibroblasts (NHGFs) compared to normal control. Ectopic miR-335-3p attenuated, whereas knockdown of miR-335-3p promoted, the fibrogenic activity of human gingival fibroblasts. Mechanically, miR-335-3p directly targeted SOS1, SMAD2/3, and CTNNB1 by canonical and noncanonical base paring. In particular, different portfolios of fibrotic markers were suppressed by silencing SOS1, SMAD2/3, or CTNNB1, respectively. Thus, our study first proposes a novel miRNA screening approach targeting a functionally related gene set and identifies miR-335-3p as a novel target for HGF treatment. Mechanically, miR-335-3p suppresses the fibrogenic activity of human gingival fibroblasts by repressing multiple core molecules in profibrotic networks. Our strategy provides a new paradigm in the treatment for HGF as well as other diseases.