The narrow-band portions of the echolocation pulses seen in cf-bats are a hypothetical substrate for target localization. This localization could be based on estimates of echo envelope amplitude and carrier frequency together with their derivatives. Evaluation of these parameters is referred to as "acoustic flow" in loose analogy to optic flow. It is assessed whether the requirements for this task may be reconciled with known principles of auditory function. For the evaluation of a single echo, this seems to be the case: auditory filter shapes provide sufficient frequency resolution; at the same time envelopes are preserved well and some noise removal is achieved. Nevertheless, should bats not be endowed with additional capabilities for noise removal, analysis of acoustic flow would be limited to favorable signal-to-noise ratios. Multiple, temporally overlapping echoes are probable in any realistic echolocation scenario. In this case, additional auditory processing steps have to be postulated, which allow simultaneous estimation of multiple carrier frequencies and reduction of demodulation distortions.