In a recent work [1], Wu et al model and utilize inter-codevectors correlations of basic VQ by context modeling and conditional entropy coding of VQ indexes to improve the performance of VQ in image compression practice substantially (about 20 percent from the previous best results). However, they need to estimate several high order conditional probabilities and find suitable weighting functions for each index. Moreover, they need to implement arithmetic coding for high order probability models.In this paper, we first classify VQ indexes into smooth and non-smooth groups by using the relative variance of each codevector. Based on the smoothness of neighboring indexes, we define three different probability models. These models describe the correlations between the current VQ index and its neighboring indexes more precisely, and adaptive arithmetic coding schemes can be applied more efficiently. Furthermore, the size of each model is guaranteed not to be greater than the square of the number of codevectors, and there is no difficulty in implementation of arithmetic coding.In one of the models, we generalize the idea of gradient match in [2] to predict the current index by use of its four known neighbors. The performance produced by our method is comparable to those in [1]. Moreover, our method is much simpler and better performance is expected in our further research.[1]. X. Wu, W. Jiang and W. Wong, “Conditional Entropy Coding of VQ Indexes for Image Compression”, Proc. of 1998 IEEE Data Compression Conference, IEEE Computer Society Press, pp.347-356, March 1998.[2]. S. Juan and C. Lee, “Entropy-Constrained Gradient-Match Vector Quantization for Image Coding”, Proc. of ICASSP 1998, pp.2665-2668.