In this paper, we propose a novel method for real estate price prediction using web new media sentiments by incorporating human searching behaivor on the web. By combining online daily news’ sentiments and Google search engine query data, we construct a web news content and online search behavior-based integrated model for real estate prediction. Besides these factors, real estate price time series data are also considered into the model in order to improve the forecasting performance. Furthermore, we make a comparison between the integrated model and the baseline model without search engine query data. Experimental results indicate that the integrated model outperforms the non-integrated model, which suggests that online user searching behavior is of great value in enhancing the prediction performance. These findings imply that the proposed integrated model is effective and feasible for real estate market prediction.