A number of vanilloids have been tested as tyrosinase inhibitors using Ligand-Based Virtual Screening (LBVS) driven by QSAR (Quantitative Structure-Activity Relationship) models as the multi-agent classification system. A total of 81 models were used to screen this family. Then, a preliminary cluster analysis of the selected chemicals was carried out based on their bioactivity to detect possible similar substructural features among these compounds and the active database used in the QSAR model construction. The compounds identified were tested in vitro to corroborate the results obtained in silico. Among them, two chemicals, isovanillin (KMapp = 1.08 mM) near to kojic acid (reference drug) in one cluster and isovanillyl alcohol (KMapp = 0.88 mM) at the same distance as hydroquinone (reference drug) in another cluster showed inhibitory activity against tyrosinase. The algorithm proposed here could result in a suitable approach for faster and more effective identification of hit and/or lead compounds with tyrosinase inhibitory activity, helping to shorten the long pipeline in the research of novel depigmenting agents to treat skin disorders.