The objectives of this research are to develop, evaluate, and compare the performance of room price forecasting models for a small hotel using data mining techniques to support efficient room pricing decisions. This study analyzed secondary data from a Property Management System, totaling 1,480 records, to evaluate the performance of Linear Regression and XGBoost Regressor models. The evaluation focused on key performance indicators: the Coefficient of Determination R², indicating the proportion of variance explained by the model, and the Mean Absolute Error (MAE), representing the average magnitude of prediction error in Baht. The results showed that the Linear Regression model achieved an R² of 0.7608 and an MAE of 1,162.27 Baht, whereas the XGBoost Regressor yielded 0.7256 and 1,112.79 Baht, respectively. Although Linear Regression exhibited a higher R², indicating a better capability to explain data variance, the XGBoost Regressor provided a lower Mean Absolute Error (MAE). In the context of pricing, minimizing the monetary margin of error is considered a more critical criterion. Consequently, this study concludes that the XGBoost Regressor is more suitable for effectively supporting dynamic pricing strategies.