Abstract: The present research aims to investigate the effect of multimodal input on the cognitive processing-route patterns in simultaneous interpreting (SI) from a corpus-assisted cognitive perspective.Based on self-supported Chinese-English parallel SI corpus, the result reveals that professional simultaneous interpreters show different patterns of cognitive processing routes under picture-aided, script-aided and non-slide-aided conditions.As to the rendering of proper names, script-aided SI yields the most memory-pairing, followed by picture-aided SI, and non-slide-aided SI yields the least memory-pairing.Regarding culture-specific non-proper items (e.g., metaphors and idioms), there is no statistically significant difference as to the impacts of multimodal input on cognitive processing routes as well as on translating strategies.It suggests that the property of the source category seems to play a vital role in the cognitive processes of SI and that the more consistent the aurally and visually presented information is, the less cognitive effort the interpreters will exert during SI.