Quorum sensing (QS) system is related to cell to cell communication as a function of population density, which regulates many physiological functions including biofilm formation and virulence gene expression. The interspecies communication is mediated by autoinducer-2 (AI-2) that is catalysed by LuxS from S-ribosylhomocysteine (RHC). QS inhibitors have emerged as a promising strategy for biofilm and virulence attenuation, which improves the potential of antimicrobial treatments by increasing microbial susceptibility. Among a wide variety of phytochemicals, many of them have been described as QS inhibitors. Driven by the promising phytochemicals clues, this study aimed to identify new active plant natural compounds against LuxS through in silico analysis, followed by in vitro validation. A representative strain for the study of LuxS inhibition, Bacillus subtilis, and a reporter strain sensitive to AI-2, Vibrio harveyi BB170, were used. Firstly, the optimization of a virtual screening protocol was conducted by applying a combination of protein-ligand docking, receptor-based virtual screening, and free energy calculations. Then, optimized virtual screening protocols were applied to screen a phytochemical database containing 3479 drug-like compounds. Based on binding energy scores and compounds cost, the most promising and selected phytochemicals were curcumin, pioglitazone hydrochloride, and 10-undecenoic acid. According to the obtained scores, 10-undecenoic acid could strongly interact with the active binding site of LuxS, while curcumin and pioglitazone hydrochloride had a slight probability compared to natural ligand (RHC). In vitro analysis corroborated the QS inhibitory activity of curcumin and 10-undecenoic acid, however, pioglitazone hydrochloride had no relevant effect. Curcumin (1.25-5 µg/mL) triggered 33-77% reduction of AI-2 accumulation and 10-undecenoic acid (12.5-50 µg/mL) reduced 36-64%. In conclusion, in silico analysis allowed the identification of LuxS antagonistic phytochemicals, revealing curcumin and 10-undecenoic acid as active QS inhibitors. Additionally, this study highlighted that although computational screening is a powerful and quick tool to identify lead compounds, in vitro validation is needed to guarantee high levels of accuracy.