Undergraduate men who held neutral positions on a variety of issues were exposed to peer-group opinions whose semantic wording was varied to present either a positive or negative context. Two replications of the procedure using statements of different contents were conducted (each sample n = 40); a 2 × 2 factorial design was used for both replications where the independent variables were direction of semantic wording (positive or negative) and direction of influence from peer ratings (for agreement or disagreement). Support was found for an interaction of the two independent variables wherein conformity increased as a function of congruity between direction of peer opinions and semantic context of stimulus-wording.