The auditory system must accurately represent sounds over a wide range of sound intensities. Spike-rate adaptation is one proposed mechanism for enhancing discrimination between salient foreground stimuli and background noise through dynamic adjustments in response sensitivity to the unfolding statistical structure of soundscapes. A critical aspect of this form of statistical learning is the interplay between feedforward and feedback pathways forming “listening loops” in the auditory brain. We performed simultaneous recordings of neuronal activity in the inferior colliculus (IC) and the primary auditory cortex (A1) in anesthetized and awake mice to investigate how statistical learning unfolds. Using high-density silicon probes, we characterized the temporal dynamics of adaptation and learning in the IC and A1. Using sound sequences of varying complexity and duration, we described the integration time properties for representation in the auditory midbrain and cortex. Our study explores the role of cortical-subcortical interactions in setting the cadences over which listening occurs; these might ultimately support high-fidelity listening performance in noisy and dynamic sound environments.