In order to solve the problem of the difficulty of pricing used sailboats in trading, this paper builds an XGBoost model based on machine learning, trains and tests it with the collected multidimensional dataset, and finally obtains a more ideal pricing model for used sailboats, and uses the model to price used sailboats now in the Hong Kong market and compare it with the actual offer. Meanwhile, we know that the hull width of the used sailboat and the gross regional product of the region where it is located are the most important factors affecting the price according to the importance of the features output from the model.