Abstract: This paper proposes an optimal steganography framework based on Fisher Information. An embedding algorithm is designed to solve the optimization problem whose objective is minimizing the Fisher Information of a steganographie system. The Fisher Information is the quadratic form of the embedding transferring probabilities. The problem of optimal embedding transferring probabilities is solved by quadratic programming. Then the optimal LSB matching algorithm, which is an instance of the optimal steganographic algorithm, is given. The experimental results show that the Fisher Information of the optimal LSB matching is smaller than that of the conventional LSB matching and the security of the optimal LSB matching is better.