The studies reported in this paper address the predictions of the social identity model of deindividuation phenomena, or SIDE (Reicher, Spears & Postmes, 1995), concerning the strategic effects of visibility to the in-group: increasing the visibility of in-group members to each other increases their ability to support each other against the out-group and hence increases the expression of those aspects of in-group identity which would attract sanctions from this out-group. In a first study, where anti-fox hunting participants were rendered accountable to pro-fox hunters, the results were the opposite of those expected: participants actually decreased their endorsement of anti-hunt disruption (which had been defined as normative for the in-group but unacceptable to the out-group) when made more visible to the in-group. These results were explained by arguing that participants perceived the intergroup relationship as participants versus experimenters rather than as anti-versus pro-fox hunters and that in-group visibility was being used to resist an experimentally imposed definition of themselves as favouring disruptive activity. This interpretation was supported in a second study where participants' accountability to the pro-fox hunters was removed, leaving them solely accountable to the experimenters, and similar results were obtained. In the final study, the relationship between student participants and staff experimenters was made the explicit topic of study. As expected students increased their endorsement of activities that are normative to students but unacceptable to staff when visible to fellow in-group members. Together these results provide further support for the SIDE model.