Trust is crucial for successful social interactions and personal well-being. Among the many mechanisms potentially favoring the evolution of trust, spillovers of social institutions on subsequent behavior have received little attention within evolutionary psychology. Although a plethora of research has investigated pro-social spillovers in economic experiments, research has thus far largely overlooked the impact of spillovers on person perception. Additionally, given that most studies on such spillovers use one-off measurement of the subsequent behavior, the stability of pro-social spillovers over time has not been investigated so far. In a three-stage laboratory experiment (n = 208), we provide evidence that pro-social spillovers occur even when perceptual cues provide an additional source of information to base decisions on. In Stage 1, participants played a series of trust games against unknown, videotaped targets to assess their baseline trust behavior against strangers, taking into account their trust perception. Subsequently, we randomly exposed them to a repeated prisoner’s dilemma in which cooperation is either favored (“C-culture”) or discouraged (“D-culture”). In the final stage – another series of 30 trust decisions against different targets – participants from the “C-culture” initially respond with higher levels of trust. This spillover effect, however, is short-lived and behavior converges back to the pre-intervention levels. Thus, our findings confirm pro-social spillovers even when participants behavior may also be influenced by perceptual cues but raise the question about their temporal stability. We complemented the experiment with an independent survey (n = 132) on the trustworthiness of the targets. Results show that these perception affects quickly gain predictive quality, suggesting the intervention effect only temporarily crowds out the trustworthiness perception as a key driver of trust behavior.