The mismatch negativity (MMN), a brain response elicited by a discriminable change in any repetitive aspect of auditory stimulation even in the absence of attention, has been widely used in both basic and clinical research during recent years. The fact that the MMN reflects the accuracy of auditory discrimination and that it can be obtained even from unattentive subjects makes it an especially attractive tool for studying various central auditory-system dysfunctions both in adults and children. In this review, we will discuss the applicability of the MMN to studies in dyslexia, which is currently thought, in the majority of the cases, to primarily result either from a dysfunction of the phonological system or a more general auditory deficit. Recent evidence indicates that the MMN enables one to determine which aspects of auditory information are deficiently processed in dyslexia. The MMN might also be helpful in the early definition of the dyslexia type, which would make it possible to start correctly-targeted training programmes before any major learning delays occur. Furthermore, the MMN holds promise of showing plastic changes in the brain of dyslexic individuals underlying the alleviation or remediation of dyslexia in the course of a successful training programme.