Many researchers consider metaphor so fundamental to psychological activity that they claim that it does not require extra cognitive effort to process. While not disputing that metaphors are natural to human cognition, we argue that a metaphor's relative ease-of-use ought not be confounded with an expectation that it prompts no extra effort. As many studies show (including those presented here), metaphors often come with costs when compared to non-figurative controls (e.g. longer processing times). However, we also argue that the extra costs associated with an apt metaphor ought to come with benefits. This analysis, based on Relevance Theory, does a good job of accounting for some overlooked psycholinguistic findings concerning metaphor processing.