Abstract: Based on data fusion the method which combines the rough sets theory and the support vector machine theory is applied in assessment of engineering quality of rock mass.Firstly,applying the rough sets theory the sample data is reduced.Removing the redundant characteristics,the concise expression can be formed expressing the relation of the factors and the engineering quality of rock mass,and new sample data be formed.Then applying the support vector machine study the new sample data to establish the support vector machine evaluation model.Through application in the example,it indicates that the model is scientific and practical,and support vector machine is a feasible method in the evaluation of tunnel rock quality.