In this letter, a new Linde-Buzo-Gray (LBG)-based image compression method using Discrete Cosine Transform (DCT) and Vector Quantization (VQ) is proposed. A gray-level image is firstly decomposed into blocks, then each block is subsequently encoded by a 2D DCT coding scheme. The dimension of vectors as the input of a generalized VQ scheme is reduced. The time of encoding by a generalized VQ is reduced with the introduction of DCT process. The experimental results demonstrate the efficiency of the proposed method. In this letter, a new Linde-Buzo-Gray (LBG) -based image compression method using Discrete Cosine Transform (DCT) and Vector Quantization (VQ) is proposed. A gray-level images are then decomposed into blocks, then each block is The dimension of vectors as the input of a generalized VQ scheme is reduced. The time of encoding by a generalized VQ is reduced with the introduction of DCT process. The experimental results demonstrate the efficiency of the proposed method.