RSKNN is an improved algorithm of KNN with better classification performance. The RSKNN algorithm is based on the theory of the variable precision rough set. The algorithm guarantees under the premise of a certain classification accuracy, effectively reduces the computation burden of the classified samples, and improves the computation efficiency and precision of classification. But the degree of dependence on attributes is very high, which can make RSKNN algorithm affected by a certain degree of precision in classification. So the use of the class subspace classification method into RSKNN algorithm can improve the classification accuracy of RSKNN. The experimental results carried out on some UCI public datasets verify the effectiveness of the proposed algorithm.