This paper presents a transform domain allocation method for watermarking of digital multimedia content; specific analytical results are obtained for audio watermarking. To solve the problem, the watermarking transmitter, channel and receiver are formally defined and watermarking is described in a detection theory framework. Describing the channel by a generic attack model and using detection theory, the challenge is to minimize the probability of detection error at the receiver. Using strong assumptions on the original and watermark signals to derive analytical results, an optimal allocation method is deduced. The allocation strategy is compared to strategies in most present systems. Moreover, the assumptions on the signals reveal to be realistic in a given implementation of the watermarking system. The allocation method is being patented for digital multimedia content.