Based on a corpus of 120 cartoons collected from the website http://cartoon.chinadaily.com.cn, this study attempts to give an account of pictorial-verbal metaphors in Chinese editorial cartoons on food safety from the perspective of Conceptual Metaphor Theory and multimodal metaphor theory. The findings are as follows: (1) Four major mode configuration patterns are observed: cross-modal mapping, mono-modal mapping, multimodal mapping and implied mapping. (2) The pictorial-verbal metaphors centre around five major target domains: unsafe food, consumer, producer/seller, supervision authority and food safety law/standard. (3) A large number of scenario metaphors are observed in the data, which can be further divided into those that are cross-culturally perceivable and those that are heavily embedded in Chinese culture. The working mechanism of scenario metaphors can be accounted for by Conceptual Blending Theory. The study helps delineate the food safety situation that we are facing in China from a metaphorical perspective. It also extends the application of multimodal metaphor theory to a Chinese context and contributes to the refinement of the theory.