Although social comparison is often considered as an automatic process, the evidence in support of this idea is weak and inconclusive. In this paper, we reexamined the question of automaticity in social comparison by testing the hypothesis that subliminal social comparison affects explicit self-evaluations. In two high-powered experiments, young women were subliminally exposed (or not) to a high standard of comparison (media images of ultra-thin women). Next, they made explicit self-evaluations of their body appearance anxiety. Using both between-participants (Experiment 1) and within-participant (Experiment 2) designs, we found converging evidence that subliminal exposure to the thin ideal increases body appearance anxiety in women. Using Bayes factors as measures of evidence, the present experiments provided substantial (Experiment 1) and very strong (Experiment 2) evidence that social comparison takes place outside awareness and affects explicit self-evaluations. The present experiments can be easily replicated using a standardized procedure (replication script) that is publicly available on the Open Science Framework. We discuss how these findings contribute to reestablish confidence in the modern view of social comparison as an automatic process.