Three experiments investigated the role of authorial intentions in metaphor comprehension. In these studies, subjects read metaphoric (e.g., "A family album is like a museum"), literal (e.g., "An art gallery is like a museum"), and anomalous (e.g., "A tortoise shell is like an art gallery") comparisons and rated their degree of meaningfulness (Experiment 1), made speeded decisions as to whether each phrase was meaningful or not (Experiment 2), or wrote out interpretations of each comparison statement (Experiment 3). The subjects were told that the comparisons were written either by famous 20th century poets or by a computer program that randomly generated the statements from a list of words. Our general hypothesis was that knowing that intentional agents (the poets) authored the different comparisons should facilitate subjects' comprehension of the metaphors. Experiment 1 showed that subjects rated both metaphoric and literal comparisons as being more meaningful in the poet condition than when these statements were supposedly written by computer. Experiment 2 demonstrated that subjects were faster in making their meaningfulness judgments for metaphors in the poet condition than in the computer context, but that subjects were also slower in rejecting anomalous comparisons when these were supposedly written by the poets. Experiment 3 indicated that subjects produced more meanings or interpretations for comparisons presumably written by poets than by computer. These results highlight the importance of implied, authorial intentions in understanding metaphorical statements. We discuss the implications of this work for psycholinguistic theories of figurative language comprehension, as well as for theories of literary interpretation.