PURPOSE: Optimal augmentative and alternative communication (AAC) systems for children with complex communication needs depend in part on child characteristics, child preferences, and features of the systems themselves. The purpose of this meta-analysis was to describe and synthesize single case design studies comparing young children's acquisition of communication skills with speech-generating devices (SGDs) and other AAC modes. METHOD: A systematic search of published and gray literature was conducted. Data related to study details, rigor, participant characteristics, design information, and outcomes were coded for each study. A random effects multilevel meta-analysis was performed using log response ratios as effect sizes. RESULTS: Nineteen single case experimental design studies with 66 participants (M age = 4.9 years) met inclusion criteria. All but one study featured requesting as the primary dependent variable. Visual analysis and meta-analysis indicated no differences between use of SGDs and picture exchange for children learning to request. Children demonstrated preferences for and learned to request more successfully with SGDs than with manual sign. Children who preferred picture exchange also learned to request more easily with picture exchange than with SGDs. CONCLUSIONS: Young children with disabilities may be able to request equally well with SGDs and picture exchange systems in structured contexts. More research is needed comparing AAC modes with diverse participants, communication functions, linguistic complexity, and learning contexts. SUPPLEMENTAL MATERIAL: https://doi.org/10.23641/asha.22111181.