With the coming of the digital era, digital imagery has become an important medium of information processing. However, there is a major problem: digital images can be forged easily. To increase the degree of credibility for digital imagery, some image tamper proofing and recovering techniques should be explored. We propose a new secure recoverable tamper proofing method. The proposed method exploits the vector quantization (VQ) technique to embed some information into the middle-frequency range of the image transformed by the discrete cosine transform (DCT), where the embedded information can be used for tamper detection and recovery. The experimental results show that the proposed method can securely and effectively detect the tamper and recover the image. Even though the image is tampered with beyond recognition, the technique can still recover the original content. Our proposed method is therefore an effective and secure tamper processing technique.