An emerging literature indicates that dispositional bias in causal attribution of social behavior is weaker for people with working-class (vs. middle-class) backgrounds. However, it is unknown whether this difference is also present in spontaneous forms of trait inference. In the current work, American undergraduates were asked to merely memorize many pairings of a target face and a trait-implying behavior. In a subsequent lexical judgment task, after each face was presented as a fixation, either an implied trait or its antonym was given as a target. As expected, participants with college-educated parents (middle class) showed a strong N400 event-related potential component to the antonym (vs. the implied trait), suggesting spontaneous trait inference during the memorization phase. In contrast, those with high-school-educated parents (working class) showed no such effect. It is important to note that the N400 spontaneous trait inference effect was associated with perceived significance of dispositions in accounting for social behaviors. (Contains 5 figures, 2 tables and 3 footnotes.)