This experiment investigated the primacy of social identification in the online social influence process with anonymity as a contextual variable. Other key variables were perceived argument strength and attitude toward abortion. Participants (N = 229) were randomly exposed to four conditions, in which they read three pro-life or pro-choice arguments from either three identifiable people (known) or three unidentifiable people (unknown). Based on the social identity model of deindividuation (SIDE), the study tested if social identification had a greater effect on attitude in unknown conditions through depersonalization. Results demonstrated that social identification predicted attitude, but anonymity did not affect the process. In other words, depersonalization was not triggered by anonymity. Social identification also affected attitude indirectly (via argument strength). The study also found presence of both group membership-based influence (driven by social identification) and informational influence (driven by argument strength). Implications of the study, especially involving anonymity, were discussed.