This paper provides results of a descriptive analysis of peer-reviewed, single subject design, intervention research on augmentative and alternative communication (AAC) for individuals with severe disabilities, from birth to 21 years, published in English between 1997 and 2003. A database of 40 studies was found that met seven specified criteria. The descriptive analysis showed that a variety of antecedent, both adult and child directed, and consequent intervention components, typically used in combination, were effective in improving communication. Most research contexts were rated as less natural. Parents, teachers, and siblings or peers were infrequently involved in intervention. When problem behavior and communication were targeted, functional communication training (FCT) was the method of choice. Treatment integrity was infrequently measured. When measured, generalization and maintenance of treatment effects were good but were reported less than half of the time. Implications for interventionists and researchers are discussed. (Contains 3 tables and 2 footnotes.)