About one-third of autistic individuals have limited functional speech, often using fewer than 30 words or simple phrases in spontaneous communication. The reasons behind this are complex and not fully understood, partly due to the limitations of traditional language and cognitive assessments in reliably measuring their abilities. Emerging evidence suggests that they may have intact receptive language skills, raising the question of why their spoken language production is impaired. Using non-invasive human neuroimaging (e.g. EEG and MEG), my research aims to examine the disrupted cognitive processes in the language comprehension-to-production pathways in non-speaking autism.