Behavioral targeting has been widely used in advertising. It recognizes potential consumers by analyzing their online behavior and displaying customized ads to individual consumer. Extant research show that a behavioral targeting ad (BTA) elicits favorable consumer reactions towards the advertised product by implying social labels that are related with consumers’ identity. Consumers tend to adjust their behaviors to match their social labels, hence showing favorable reactions towards the product in a BTA. This research investigated and revealed that a boundary condition for the effectiveness of a BTA—the valence of the implied social label from a BTA. Specifically, we found that participants show more favorable reactions towards a BTA than a traditional ad only when it implied positive versus negative social labels. This is because it is human nature to associate oneself with positive social labels but to avoid links with negative labels. Hence, BTAs will perform better than traditional ads when the implied social labels are positively-valenced and will consequently invite favorable consumer reactions. However, this effect will disappear when the implied social labels are negative, driven by people’s motivation to avoid negative social labels. Our findings advance our understanding of BTA and offers practical implications for marketers.