Involvement as a moderator of message sidedness effects has produced inconclusive and contradictory findings. This study examines these contradictions by investigating the conditions in which involvement moderates the persuasiveness of two-sided advertising. Based on meta-analytic results and using two experimental studies, the findings show that high involved consumers are influenced by the negativity of the message, while low involved consumers are influenced by the amount of information. Furthermore, the influence of negativity on high involved consumers is conditional when there are low degrees of negativity in the message. In comparison, varying the degree of negativity does not affect low involved consumers.