Traditionally, communication interventionists focused on teaching a beginning repertoire of communicative behavior, once learners with severe to profound disabilities had emitted intentional behavior. Increasingly, interventionists are recognizing that valuable opportunities may be lost if intervention does not begin at an earlier point. In part, intervention strategies at increasingly earlier points have resulted from a prevailing change from semantically-focused intervention logic to pragmatic, interaction-focused intervention logic. At the same time that intervention content has increasingly focused on pragmatics, there has been a wealth of information addressing the social functions served by repertoires of simple idiosyncratic (as well as socially unacceptable) behavior. The increasing availability of augmentative and alternative communicative options has provided an extensive array of motorically simple strategies to exert significant control and influence over one's environment. We have long since passed the need to demonstrate that persons with severe disabilities can be taught a repertoire of communicative functions. However, we have not been as successful in demonstrating that the communicative behavior taught is well maintained solely in the presence of natural maintaining contingencies. Nor have we adequately demonstrated that established repertoires are sufficiently generalized. Most recently, interventionists have begun to focus on more efficient strategies to use in the selection of the most critical teaching instances to teach a new communicative response. Additionally, interventionists are considering response efficiency as an important variable in determining the likelihood that a learner will choose to emit members of his or her communicative repertoire. There appears to be a consensus among those who currently serve individuals with severe disabilities that inclusion represents an attainable objective for students with even the most severe disabilities. Unfortunately, it is not clear that either special or regular educators are being adequately prepared to accomplish included placements. There remains a significant need to recognize those aspects of best practice which must be further explored in regular education settings. What once were considered best practice methods may not meet the test of social validity and be considered best practices in regular classrooms. The vast majority of intervention research has selected a fairly narrow communicative form or function to teach. Increasingly, information on maintenance and generalization is considered. However, often the periods sampled postacquisition are very modest. Among the plethora of available communication intervention curricula, there are virtually none that have taken a learner from a point of engaging in no intentional communicative behavior to the establishment of an effusive repertoire of communicative functions and corresponding vocabulary.(ABSTRACT TRUNCATED AT 400 WORDS)