Abstract: Mineral price forecasting is the necessary premise of mining projects investment decision and important basis of risk evaluation. Regarding molybdenum ore as the subject, we set up a mineral price forecasting model fitting for the mining conversion and expansion with least squares method by adopting the following research technologies, such as systematic theory, empirical study, qualitative analysis and quantitative calculation. Based on the analysis of the historic mineral prices, we use the polynomial fitting way to predict the price of minerals in recent years.