In this paper, we proposed an adaptive secure client-side digital audio watermarking by choosing an optimal embedding key from numerous available embedding keys for the encryption and joint decryption and watermarking procedure of the specific content. In the proposed scheme, the embedding key will determine not only the client-specific watermark content to be embedded, but also the watermarking embedding position where the watermark will be placed. It is obvious that different region of the host signal featured different characteristics, so where to embed the watermark will impact the equilibrium between two basic requirements, known as inaudibility and robustness, of watermarking development. Consequently, how to choose the embedding key for the particular content to be distributed could be of crucial significance. Simulation results show that the watermark in this scheme can survive robustly after being subjected to various common audio attacks, such as filtering, resample and so on. © 2011 Springer-Verlag.