An exploration of metaphor and literal comprehension is outlined within the framework of an episodic verification task that allows a criterion of comprehension more accurate than standard comprehension paradigms. Strategies used by Ss in metaphor comprehension in different contextual settings and Stroop-like effects between literal and metaphorical meanings of a sentence were also examined in three experiments. Experiment 1 examined the verification of literal and metaphorical sentences with respect to a previously learned character description. Undergraduate Ss (N = 23) heard one of two stories describing different characters and made true/false judgments about literal and metaphorical statements relating to the character. Within-Ss factors were type of sentence and truth value, whereas between-Ss variables were character and list (two lists of sentences were constructed). In experiment 2, the same stimuli and materials were used, but the verification list contained only literal or only metaphorical sentences (N = 24). Experiment 3 explored within-list contextual effects using semantically unrelated target sentences (N = 64). In experiments 1 and 2, verification was slower for metaphors than for literals and literals only took longer on mixed lists than on single-type lists. In experiment 3, a context sensitivity to within-list composition was found for both sentence types. Overall effects indicate Ss' capability of strategically modifying the range of meanings activated by target sentences. (Copyright 1992, Sociological Abstracts, Inc., all rights reserved.) (LLBA 1992, vol. 26, n. 1)