The auditory perception is also an important sensing modality for intelligent robots. It demonstrates superiority especially in situations where the objects are visually indistinguishable. To recognize visually identical bottles with different materials in them, it is possible to interact with bottles. And different materials can make different sounds, and thus the target bottle can be recognized. In this chapter, an active auditory recognition task is proposed, in which the robot actively interacts with bottles with different actions generating different sounds and then recognizes the content in bottles. Additionally, the active auditory exploration strategy is established based on the reinforcement learning. Case studies are conducted to demonstrate that the learned exploration strategy is effective in generating actions for the robot to actively interact with the bottles to generate sounds, and satisfying performance has been achieved for the auditory recognition task.