Objective: This study assessed a new Speech Generating Device (SGD) for supporting request responses in five participants with intellectual and multiple disabilities. Methods: The technology involved a smartphone, a series of mini objects or of cardboard chips with pictures, and special software. When the participants placed one of the mini objects or chips with pictures (that they carried at their waist) against the smartphone (that they had at their chest), the smartphone emitted a verbal request concerning the activity indicated by the mini object or chip. The study was carried out according to a non-concurrent multiple baseline design across participants using sessions of 20 min. Results: During the baseline, the participants' mean frequencies of requests were zero or close to zero. Following the 11-15 intervention sessions, all participants were successful in making requests (i.e. their mean request frequencies were between about five and 12 per session). Three participants showed clear preferences (i.e. more requests) for some of the activities. The other two participants were rather varied in their requests within and across sessions. Conclusions: The new SGD seems very useful for people with multiple disabilities who are blind or have poor control of their fine motor responses.