Abstract: In view of the existing coal seam gas content prediction data recording and shortcomings of processing equipment, the paper studied the reason for error induced by the existing curve fitting method. After a large number of mathematical modeling, through two kinds of curve fitting algorithm optimization, and considering the demand of test, the design of the upper and lower units of an intelligent coal seam gas content determination system was completed. The lower unit has plural groups of storage, time setting, desorption rate curve display, logical deduction and loss fitting. The upper unit based on VC++ program design has functions such as interface development, database inquiry and fitting algorithm processing. Results show that the intelligent system has greatly improved the operation and accuracy of data fitting.