This study aims to screen and validate AURKA inhibitors to provide lead compounds and theoretical foundations for targeted therapy of neuroblastoma (NB). Through computer-aided drug screening, 11 compounds effectively binding to AURKA were selected from the YaTCM database, and their toxicity was predicted using admetSAR. Subsequently, molecular dynamics (MD) simulations were employed to evaluate the binding affinity and complex stability of the compounds with AURKA, leading to the identification of four preferred compounds (Erylatissin B, (+)-khellactone, Brazilin, and hematoxylin). Further steered molecular dynamics (SMD) and umbrella sampling (US) simulations were conducted to calculate the dissociation energy, confirming their binding strength with AURKA. In vitro experiments demonstrated that Brazilin significantly inhibited proliferation, migration, and induced apoptosis in SK-N-BE (2) cells, while also suppressing AURKA protein expression and its interaction with N-Myc. In vivo experiments showed that Brazilin markedly inhibited tumor growth in a mouse NB model. The findings indicate that Brazilin, by targeting AURKA, exhibits potential anti-NB activity, offering a new candidate compound and theoretical support for NB treatment.