Abstract: Prediction of rock burst is a complex system problem of uncertainty.For the sake of accurate evaluation of the estate of rock burst,providing a scientific basis for the prevention of rock burst,this paper brings forward a prediction method for rock burst based on the fractal interpolation theory.First,three key factors are chosen to construct the model.In addition,because of few standard samples,forty standard samples are produced by the way of stochastic interpolation on the basis of classification standard.Second,by calculating the fractal dimension of single index of rock burst,the fractal dimension exponent evaluation of the index of single rock burst is confirmed with the maximum likelihood classification principle.Third,using the weighted summation value method,the level of the colligate estimation of multi-index of rock burst is calculated and the fractal interpolation evaluation model is established according to the relationship between the value of comprehensive evaluation of the samples and the experience level.Finally,some samples are classified based on the model.The classification results are consistent with those from other methods.Results show that evaluation of the fractal interpolation model is reasonable and objective,and it has better reliability and stability to provide a new research method and idea for rock burst prediction.