<p>Although some anti-bullying programs stimulate defending victims, it is unclear whether defending effectively supports victims. One problem hampering research is that measures of defending not only reflect defending by <i>pure defenders</i> who do not have any other role in bullying. They also reflect defending by (1) <i>low profile defenders</i> who also are victims or outsiders and who can be assumed to lack social dominance and to only defend indirectly (comforting victims rather than confronting bullies); (2) <i>antisocial defenders</i> who also are <i>probullies</i> (bullies, assistants, and reinforcers) and who may lack the prosocial attitude that underlies defending actual victims; and (3) <i>ambivalent defenders</i> who simultaneously are low profile and antisocial defenders. Because the severity of such a contamination of defending measures depends on the prevalence of defender subtypes, we used peer nominations to classify the defenders among 6554 Dutch adolescents (<i>M</i>_age = 13.3, <i>SD</i> = .5; 48% boys) into defender subtypes and in terms of their defending strategies (indirect, direct, or indirect/direct). To test the assumptions mentioned above, we compared the defender types to each other and to victims, outsiders, and probullies in terms of social dominance and, as a proxy for a prosocial attitude, social preference. Most defenders either were low profile defenders who defended indirectly and lacked social dominance, or antisocial or ambivalent defenders who lacked social preference. Accordingly, only a minority of defenders may effectively support victimized peers. We encourage future researchers to differentiate between defender subtypes when examining defending effectiveness and when stimulating defending.</p>