Kriging interpolation is a widely used low-pass filter interpolation method, but it cannot reconstruct the high-frequency, low-frequency and partial information of original information. Fractal interpolation using self-similarity, which can retain original information, overcomes the limitations of Kriging interpolation low-pass filters, thereby improving the interpolation accuracy. On the basis of the traditional fractal interpolation algorithm and in combination with geological spatial information, this paper introduces a four-dimensional space fractal interpolation algorithm suitable for ore grade estimation. The interpolation algorithm is applied to molybdenum ore grade estimation and then compared with the Kriging interpolation algorithm. The results show that the interpolation algorithm is superior to the Kriging interpolation algorithm.