The paper compares different approaches to researching metaphors, highlighting their limits and possibilities. Psycholinguistic methods, for instance, are considered valuable for their insight into the comprehension of metaphoric expressions during online tasks and especially into the distinctive features of online metaphor comprehension and literal language understanding. Reaction time studies, which enabled a first substantiation of existing theories of figurative language comprehension, are nonetheless shown to be not entirely capable of detecting differences occurring within the same or comparable time frame. Electrophysiological methods, on the other hand, are proved to be more effective, as event-related potential studies report differences between metaphoric and literal language comprehension that occur as early as 200 ms (P200) after the presentation of the critical word. To complete this survey, source localization methods are indicated as efficacious in the identification of the neural generators of these components.