The Ease of Language Understanding model (Rönnberg, 2003; Rönnberg et al., 2008; Rönnberg et al., 2011) predicts that speech understanding in adverse, mismatching noise conditions is dependent on explicit processing resources such as working memory capacity (WMC). This presentation will examine this prediction by addressing some data on (1) how brainstem responses are affected by WMC, (2) how cortical networks of speech understanding in noise are modulated by WMC, and (3) how an inhibition sub-process of WMC determines episodic long-term memory for spoken discourse masked by speech.