Since the famous publication of Baron and Kenny's (1986) article about 20 years ago, in social psychology research the terms mediators and moderators are used more and more frequently (MacKinnon, Fairchild and Fritz, 2007). A debate on conceptual framework and appropriate data-analytic procedures has been raised. First of all, this article intends to delineate conceptual distinctions between the two terms in order to emphasize different explanation of theoretical models and application to evaluation research of psychosocial programs. Secondly, recent developments on data-analytic procedures in moderation and mediation analysis are presented; main strengths and limitation of moderation and mediation analysis in social psychology are considered.