This paper is a follow-up study on the issue of L2 inner speech as it manifests in mental rehearsal among advanced L2 learners. The purpose of the study was to find out to wliat extent advanced L2 learners experience inner speech as mental rehearsal and to identify some of the characteristics andfunctions ofsuch inner speech. Results show that advanced L2 learners experience inner speech in the second language to a great extent and that the frequency ofL2 inner speech increases with proficiency. Advanced L2 learners, however, report using less inner speech than lower level learners for certain aspects of rehearsal, such as planning texts, self-and other-evaluation, storage and retrieval, self-instruction, and language play. It is argued that inner speech in the L2 is a developmental phenomenon associated with spontaneous rehearsal in the early stages ofL2 acquisition and with verbal thinking in the more advanced stages.The problem of inner speech is central in Vygotsky's (1978Vygotsky's ( , 1986 theory. As Vygotsky's and his followers' ideas have become globally recognized, inner speech has been acknowledged as a major phenomenon associated primarily with the LI. In the field of second language acquisition (SLA), however, inner speech continues to be practically uncharted territory. 1 At least two reasons can be cited to account for this lack of attention on the part of SLA researchers. First, it is possible that, for researchers working within the current prevailing SLA paradigm, Vygotsky's theories and the problems associated with them appear to be irrelevant and/or irreconcilable. A problem like inner speech, which suggests a view of mind as a predominantly sociocultural product, framed as it was within the dialectics of historical materialism in Vygotsky's writings, 2 does not seem to "fit the facts" (as Kuhn, 1970, p. 141, would put it) of SLA mainstream theory, which is ultimately concerned with the psychological mechanisms that underlie the acquisition of L2 properties as a process situated mainly in the learner's head. 3 Second, there is the problem of method. Because inner speech is covert language behavior, it is inaccessible to direct methods of observation. Vygotsky (1986) himself recognized that "the area of inner speech is one of the most difficult to investigate" (p. 226).His way of breaking the inaccessibility of inner speech was the "genetic method of experimentation" (p. 226). This method, which came to be known as "the Vygotsky method of studying inner speech" (Ushakova, 1994, p. 137), consisted of approaching inner speech through the observation and analysis of egocentric speech, in the assumption that egocentric speech is the vocalized transition between social external speech and inner covert speech. As SLA research becomes more open to nontraditional theoretical and methodological approaches, however, topics like Issues in Applied Linguistics