A new goal-systems model is proposed to help explain when individuals will protect themselves against the risks inherent to social connection. This model assumes that people satisfy the goal to feel included in safe social connections—connections where they are valued and protected rather than at risk of being harmed—by devaluing rejecting friends, trusting in expectancy–consistent relationships, and avoiding infectious strangers. In the hypothesized goal system, frustrating the fundamental goal to feel safe in social connection sensitizes regulatory systems that afford safety from the risk of being interpersonally rejected (i.e., the risk-regulation system), existentially uncertain (i.e., the social-safety system), or physically infected (i.e., the behavioral-immune system). Conversely, fulfilling the fundamental goal to feel safe in social connection desensitizes these self-protective systems. A 3-week experimental daily diary study (N = 555) tested the model hypotheses. We intervened to fulfill the goal to feel safe in social connection by repeatedly conditioning experimental participants to associate their romantic partners with highly positive, approachable words and images. We then tracked how vigilantly experimental versus control participants protected themselves when they encountered social rejection, unexpected behavior, or contagious illness in everyday life. Multilevel analyses revealed that the intervention lessoned self-protective defenses against each of these risks for participants who ordinarily felt most vulnerable to them. The findings provide the first evidence that the fundamental goal to feel safe in social connection can co-opt the risk-regulation, social-safety, and behavioral-immune systems as independent means for its pursuit.