Metaphors are considered as instruments crucial for persuasion. However, while many studies and works have focused on their emotive, communicative, and persuasive effects, the argumentative dimension that represents the core of their “persuasiveness” is almost neglected. This paper addresses the problem of explaining how metaphors can communicate arguments, and how it is possible to reconstruct and justify them. To this purpose, a distinction is drawn between the arguments that are communicated metaphorically and interpreted based on relevance considerations, and the ones that are triggered implicitly by the use of a metaphorical expression. In both cases, metaphorical arguments are reconstructed through different patterns of argument, called argumentation schemes (Walton, Reed and Macagno 2008). However, while the purpose of a metaphorical sequence of discourse (called metaphorical move) can guide and justify the reconstruction of the argument that can sufficiently support the intended conclusion in a persuasive move, a more complex analysis is needed for analyzing the additional inferences that a metaphorical move can trigger. These inferences are claimed to represent part of the connotation of the metaphorical expression and can be captured through its most frequent collocations, determinable using some tools of the corpus linguistics.